Out-of-hospital cardiac arrest (OHCA) is an important health care issue. In a systematic review of over 60 studies, the incidence and outcome of OHCA varied greatly. The mean incidence of adult OHCA globally was estimated to be 55 per 100 000 person-years and the survival-to-discharge rate ranged from 2% to 11%.1 Lack of uniform reporting practices is believed to be one of the reasons for this variability. In Hong Kong (HK), there have been at least four studies since 1995 reporting the outcome of OHCA.2 3 4 5 All were either a single-centre study or recruited cases from a particular district of HK. The survival-to-discharge rate was between 0.5% and 3%.2 3 4 5 There was no estimate of the incidence of OHCA and the reporting style also varied. This study is the first local territory-wide investigation of OHCA. The objective was to evaluate the epidemiology and outcomes of OHCA in HK. It is hoped that lessons can be drawn from the findings to improve OHCA survival locally.

Hong Kong covers approximately 1100 km² with a population of about 7.3 million. In 2014, it attracted nearly 61 million overseas visitors.6 The emergency medical service (EMS) is provided by the Hong Kong Fire Services Department (FSD) and is a one-tier system. The EMS dispatchers do not give advice about cardiopulmonary resuscitation (CPR) even when OHCA is reported. Transport is primarily by ground vehicles. The level of training of the ambulance crew is similar to that of an Emergency Medical Technician–intermediate level. In cases of OHCA, in addition to basic life support and defibrillation, emergency crews can provide intravenous fluids and airway management via a laryngeal mask airway. No mechanical CPR device is used. There is no protocol for pre-hospital targeted temperature management. The ambulance crews are not allowed to withhold CPR for OHCA patients prior to arrival at hospital unless it is clearly inappropriate, eg with decapitation or decomposition. Such patients are directly transferred from the scene to the public mortuary. Advance directives or do-not-resuscitate orders are uncommon in HK. Emergent coronary reperfusion and therapeutic hypothermia for OHCA patients who survive to hospital admission is performed in selected hospitals only.

This study was a retrospective analysis of the OHCA database compiled prospectively by EMS personnel from 1 August 2012 to 31 July 2013. Data included characteristics of the patient (age and gender), the cardiac arrest (location, whether the event was witnessed, whether there was bystander CPR or defibrillation with an automated external defibrillator [AED], and initial electrocardiogram [ECG] rhythm detected by the EMS), and the EMS response time (time of recognition, call receipt, arrival at patient’s side, arrival at the emergency department [ED], delivery of bystander CPR, delivery of CPR by EMS, and first defibrillation). Outcome data (whether there was return of spontaneous circulation before arrival at the ED; whether resuscitation was conducted in the ED; whether the patient survived to hospital admission, at 30 days, or to hospital discharge; and neurological status on hospital discharge) were collected by the Medical Director of the FSD from hospital records on the electronic database of the Hospital Authority. Patients of all ages were included in this study. Exclusion criteria were OHCAs caused by trauma or those victims directly transferred to the public mortuary from scene by EMS personnel, and patients not using a ground ambulance. Data were reported according to the Utstein style.7 Descriptive statistics were used to describe the patients, OHCA, and timeliness of the EMS. Patients who survived at 30 days were compared with non-survivors in terms of patient characteristics, OHCA characteristics, and the EMS response time. Significance testing by Chi squared test and Mann-Whitney U test was done for categorical and continuous variables, respectively. A P value of <0.05 was considered statistically significant. To identify predictors of 30-day survival, multivariate logistic regression with backward stepwise selection was performed. Covariates used in the analysis were age, gender, arrest location, witnessed arrest, bystander CPR, bystander AED, initial shockable rhythm, recognition to activation interval, time taken to reach the patient, time to first defibrillation, and time to ED arrival. Findings are reported as odds ratios with 95% confidence intervals. Model calibration was performed by the Hosmer-Lemeshow goodness-of-fit test. Statistical analysis was performed by the Statistical Package for the Social Sciences (Windows version 23.0; SPSS Inc, Chicago [IL], US). This study was approved by the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster (Reference number: UW 15-599), with the requirement of patient informed consent waived because of its retrospective nature.

The EMS attended 5154 OHCAs during the study period. This corresponded to 72 arrests per 100 000 persons (based on the Hong Kong Monthly Digest of Statistics published by the Government).8 The median age of patients was 80 years. Males outnumbered females by 12% and approximately 60% of OHCAs were unwitnessed. The arrests most commonly occurred at the patient’s home. Bystander CPR and AED were performed in 28.8% and 1.4% of cases, respectively. Asystole was the initial ECG rhythm on site in approximately 80% of patients. Ventricular fibrillation (VF) or ventricular tachycardia (VT) was evident in only 8.7% of patients (Table 1).

The median response time of the EMS was 9 minutes. The median time to administration of first defibrillation was 12 minutes (Table 2). The time to coronary reperfusion was not available from the database.

Of the 5154 cases, 162 (3.1%) had return of spontaneous circulation (spontaneous circulation with a palpable pulse) on hospital arrival; 788 (15.3%) patients survived to hospital admission; and 121 (2.3%) were still alive at 30 days or survived to hospital discharge and 78 (1.5%) had a good neurological outcome (defined as cerebral performance category score ≤2) [Table 3]. Survivors and non-survivors were compared, and statistically significant differences were found regarding age, gender, location of arrest, witness status, bystander defibrillation, initial ECG rhythm, recognition-activation interval, time from call receipt to patient’s side, and time to first defibrillation (Table 4). Further analysis with multivariate regression analysis identified location of arrest, initial ECG rhythm, and time from call to first defibrillation as independent predictors of survival at 30 days (Table 5). The Hosmer-Lemeshow test yielded a P value of 0.82.

This is the first territory-wide study that reports the epidemiology and outcomes of OHCA in HK. The incidence of 72 arrests per 100 000 persons is higher than the global average of 55 per 100 000 person-years. An ageing population, of whom nearly 15% are older than 65 years, probably contributes to this difference.9 The significance of an ageing population in HK is also reflected by the high median age of 80 years. The OHCA patients in HK are older than those in Japan and are probably the oldest in Asia.10 Nearly 30% of OHCAs occurred in a home for the aged (HFA) or nursing home for the elderly. Whether this is a contributing factor to the relatively low survival rate revealed by this study is unclear. In the 1990s, it was suggested that the prognosis of OHCA in nursing home residents was dismal.11 More recent studies, however, have revealed contradictory results. A Danish study published in 2014 found similar survival rates between nursing home and non–nursing home victims of OHCA.12 On the contrary, a study in Melbourne showed that survival following OHCA that occurred in residential aged care facilities was poorer than that occurred elsewhere.13 In HK where 30% of OHCAs occur in a HFA, the prognosis of local HFA residents with cardiac arrest warrants further investigation. The bystander CPR rate of 28.8% revealed by this study has almost doubled from 15.6% in 1999.3 There is still room for improvement, however. Community-based CPR education programmes are ideal as public knowledge and attitudes towards CPR are poor.14 The rate of defibrillation by bystanders was very low. This is due to the absence of a comprehensive programme of public access to defibrillation in HK. This study also revealed a drop in the percentage of VF/VT as the initial rhythm from 14% in 1999 to 8.7% in this study.15 This decline is compatible with the findings reported elsewhere.16 It remains to be determined whether this is due to the drugs used in treating coronary artery disease.

The local EMS response time (time between call receipt and reaching the patient) of 9 minutes is in the mid-range compared with major Asian cities.10 The call–to–first defibrillation interval of 12 minutes is long. This is probably related to the AED protocol at that time. In cases of unwitnessed arrest, the ambulance crews were required to perform 2 minutes of CPR before rhythm detection. With a time lag of 12 minutes, the chance of successful defibrillation by the EMS would be low. Of note, HK is a highly urbanised and densely populated metropolis. Improving the timeliness of the EMS may be difficult because of traffic congestion. Promotion of public access to defibrillation with a consequent reduced time between OHCA and delivery of the first defibrillation in appropriate patients may improve the success of resuscitation for OHCA victims. Moreover, since 2015, the AED protocol used by the EMS has been amended to immediate rhythm detection even for unwitnessed arrests.

The survival rate of 2.3% in HK is low compared with western countries.1 It is also at the lower range of survival rate among the major Asian cities.10 A high median age of OHCA patients in HK may be a factor, but it is not the only cause. A US study of over 100 000 patients found that age alone was not a good predictor of outcome.17 In another Swedish study, it was found that in the presence of favourable resuscitation factors—eg witnessed arrest, cardiac aetiology with VF as the initial rhythm—a survival rate of 10% could be achieved even in those aged 90 years or above.18 Another explanation for the low survival rate in HK is the low percentage of VF/VT as the initial rhythm. Patients with OHCA with a rhythm other than VF have been shown to have a worse prognosis.15 The low rate of VF/VT may in turn be influenced by the low bystander CPR rate and a large proportion of OHCAs being unwitnessed. Another factor contributing to the low survival rate is that HK EMS personnel have to treat almost every patient with OHCA, even when it is clear they are deceased or have been dead for some time. If a protocol permitted EMS personnel to withhold CPR from the obviously dead, the calculated survival rate would have been higher than 2.3%. Although there was a significant difference between survivors and non-survivors at 30 days or on hospital discharge in terms of age, gender, location of arrest, witness status, bystander defibrillation, initial ECG rhythm, recognition-activation interval, time from call receipt to patient’s side, and call–to–first defibrillation interval, only three were independent predictors of 30-day survival. Arrest occurring en-route to hospital is the most powerful one. This is not unexpected as immediate life support measures can be instituted by the EMS. Collapse in public areas also favours survival as the arrest is more likely to be witnessed and resuscitative measures implemented sooner. The finding that the presence of a shockable initial rhythm and prompt defibrillation favours survival concurs with multiple studies of the relationship between VF/VT in OHCA and survival.19 Bystander CPR was not a significant predictor of survival in this study, contrary to the common belief that bystander CPR improves survival. This raises the concern that in those 28.8% of OHCA victims who received bystander CPR, the quality of CPR was suboptimal. This warrants a separate investigation.

The primary limitation of this study was that the analysis was based on data collected for 12 months only and cases not attended by the EMS in the first instance were not included. The latter included an unknown number of patients with advance directives or do-not-resuscitate orders. Data from a longer period of time and the inclusion of cases taken to hospital by means other than ground ambulances or agencies like St John Ambulance should give more accurate results. The analysis of survival was also limited by the lack of data on post-resuscitation care that could have produced a more accurate estimation of the predictors of survival. Furthermore, because of the methodological limitation in the logistic model for 30-day survival predictors, only 638 cases were included for the calculation. Caution is recommended when interpreting survival predictors.

In the presence of an ageing population, OHCA is becoming a significant health care problem in HK. The survival rate of OHCA patients is low. Targeted measures such as implementation of a territory-wide public access defibrillation programme and community-based CPR education and training are needed to improve the chance of survival following OHCA.

The authors would like to thank the Hong Kong Fire Services Department for assistance in data collection. The authors thank Mr Reynold Leung for providing technical support to the project, including data cleaning, statistical analysis, and drawing figures and tables for manuscript improvement.

All authors have disclosed no conflicts of interest.

1. Berdowski J, Berg RA, Tijssen JG, Koster RW. Global incidences of out-of-hospital cardiac arrest and survival rates: Systematic review of 67 prospective studies. Resuscitation 2010;81:1479-87. Crossref

2. Wong TW, Yeung KC. Out-of-hospital cardiac arrest: two and a half years experience of an accident and emergency department in Hong Kong. J Accid Emerg Med 1995;12:34-9. Crossref

3. Leung LP, Wong TW, Tong HK, Lo CB, Kan PG. Out-of-hospital cardiac arrest in Hong Kong. Prehosp Emerg Care 2001;5:308-11. Crossref

4. Wai AK, Cameron P, Cheung CK, Mak P, Rainer TM. Out-of-hospital cardiac arrest in a teaching hospital in Hong Kong: descriptive study using the Utstein style. Hong Kong J Emerg Med 2005;12:148-55.

5. Lau CL, Lai JC, Hung CY, Kam CW. Outcome of out-of-hospital cardiac arrest in a regional hospital in Hong Kong. Hong Kong J Emerg Med 2005;12:224-7.

6. Hong Kong—the facts. Government of Hong Kong Special Administrative Region. Available from: http://www.gov.hk/en/about/abouthk/facts.htm. Accessed Jan 2016.

7. Perkins GD, Jacobs IG, Nadkarni VM, et al. Cardiac arrest and cardiopulmonary resuscitation outcome reports: update of the Utstein Resuscitation Registry Templates for Out-of-Hospital Cardiac Arrest: a statement for healthcare professionals from a task force of the International Liaison Committee on Resuscitation (American Heart Association, European Resuscitation Council, Australian and New Zealand Council on Resuscitation, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Council of Southern Africa, Resuscitation Council of Asia); and the American Heart Association Emergency Cardiovascular Care Committee and the Council on Cardiopulmonary, Critical Care, Perioperative and Resuscitation. Circulation 2015;131:1286-300. Crossref

8. Hong Kong Monthly Digest of Statistics. Government of Hong Kong Special Administrative Region. Available from: http://www.statistics.gov.hk/pub/B10100022013MM12B0100.pdf. Accessed Jan 2016.

9. Census and Statistics Department. Government of Hong Kong Special Administrative Region. Available from: http://www.censtatd.gov.hk/home/. Accessed Jan 2016.

10. Ong ME, Shin SD, De Souza NN, et al. Outcomes for out-of-hospital cardiac arrests across 7 countries in Asia: The Pan Asian Resuscitation Outcomes Study (PAROS). Resuscitation 2015;96:100-8. Crossref

11. Benkendorf R, Swor RA, Jackson R, Rivera-Rivera EJ, Demrick A. Outcomes of cardiac arrest in the nursing home: destiny or futility? Prehosp Emerg Care 1997;1:68-72. Crossref

12. Søholm H, Bro-Jeppesen J, Lippert FK, et al. Resuscitation of patients suffering from sudden cardiac arrests in nursing homes is not futile. Resuscitation 2014;85:369-75. Crossref

13. Deasy C, Bray JE, Smith K, et al. Resuscitation of out-of-hospital cardiac arrests in residential aged care facilities in Melbourne, Australia. Resuscitation 2012;83:58-62. Crossref

14. Chair SY, Hung MS, Lui JC, Lee DT, Shiu IY, Choi KC. Public knowledge and attitudes towards cardiopulmonary resuscitation in Hong Kong: telephone survey. Hong Kong Med J 2014;20:126-33.

15. Fan KL, Leung LP. Prognosis of patients with ventricular fibrillation in out-of-hospital cardiac arrest in Hong Kong: prospective study. Hong Kong Med J 2002;8:318-21.

16. Cobb LA, Fahrenbruch CE, Olsufka M, Copass MK. Changing incidence of out-of-hospital ventricular fibrillation, 1980-2000. JAMA 2002;288:3008-13. Crossref

17. Andersen LW, Bivens MJ, Giberson T, et al. The relationship between age and outcome in out-of-hospital cardiac arrest patients. Resuscitation 2015;94:49-54. Crossref

18. Libungan B, Lindqvist J, Strömsöe A, et al. Out-of-hospital cardiac arrest in the elderly: A large-scale population-based study. Resuscitation 2015;94:28-32. Crossref

19. Sasson C, Rogers MM, Dahl J, Kellermann AL. Predictors of survival from out-of-hospital cardiac arrest: a systematic review and meta-analysis. Circ Cardiovasc Qual Outcomes 2010;3:63-81. Crossref

In this current era when personal privacy is highly regarded, it may seem counterintuitive for patients to prefer the presence of a third party during an intimate physical examination (IPE). The presence of a medical chaperone, however, may offer comfort and psychological support for patients in protecting them against indecent behaviour, and is now an established good practice.1 Conversely, in an increasingly litigious society, a chaperone may serve as a witness to protect doctors against false allegations.2 Usually IPEs involve per-rectal, genital, or breast examinations, but may also include physical contact with any other part of the patient’s body.

The majority of reports about patients’ perceptions of a medical chaperone have focused on female patients3 4 5 6 7; those of male patients are less well studied.8 9 10 A potential discrepancy may exist between the two groups. Santen et al10 reported that in the US, a great majority of male patients (88%) did not care about the presence of a chaperone, while half of the female patients would prefer to be chaperoned when examined by a male physician, and a quarter when examined by a female doctor. Moreover, previous studies have been largely based on patient populations in English-speaking countries. It has been suggested that the presence of a chaperone is a ‘western concept’ that may receive a different degree of emphasis or acceptance in other parts of the world.11 To the best of our knowledge, there has been no related report on patients in Hong Kong. In this study, we investigated patients’ attitudes towards and experiences with the presence of a medical chaperone, and in particular, the relevant impact of patient and physician gender.

We conducted a cross-sectional questionnaire survey of patients waiting for their consultations at the Accident and Emergency Department and the surgical out-patient clinic of a public teaching hospital over a period of 2 weeks from late February to early March 2015. Participants were recruited irrespective of age, gender or ethnicity, but were excluded if they were under 18 years of age, or did not understand English, Cantonese, or Mandarin. Participants from the surgical out-patient clinic included both new and follow-up patients from various subspecialties. Informed patient consent and approval from the Institutional Review Board of our institution were obtained.

The questionnaire, written in both English and Chinese, comprised eight questions on demographics, and 19 questions on previous experiences with IPE, preferences for the presence of a chaperone, influence of gender of the examining physician, and the underlying reasons for their preferences. Participants selected their answers from predetermined options. In this study, IPE was defined as breast, pelvic, genital, and/or per-rectal examination. This was explained to the participants at the beginning of the survey. Guidance on the survey was provided by an investigator if requested by the participant.

Data were analysed using the Statistical Package for the Social Sciences (Windows version 22.0; SPSS Inc, Chicago [IL], US). Descriptive analyses were performed on demographic data and participant responses. Patient ages were grouped into age ranges (<40, 40-59, and ≥60) for analysis. The independent variables were patient gender and other demographic variables (eg age, education level). The dependent variables were preference for the examining doctor’s gender, presence of a chaperone with a male examining doctor, gender of chaperone with a male examining doctor, presence of chaperone with a female examining doctor, and gender of chaperone with a female examining doctor. Bivariate analyses using the Chi squared test were performed to compare the various independent variables against the dependent variables. Other variables based on participant responses—such as reasons behind chaperone preferences, previous experience of IPE, and tendency for litigation if felt harassed—were analysed against other variables as appropriate. Statistical significance was set at a probability level of 0.05.

Of a convenience sample of 183 patients, 33 declined to participate and 150 patients were recruited. Of these patients, 83 (55.3%) were from the Accident and Emergency Department and 67 (44.7%) were from surgical out-patient clinics. Recruitment was done in both locations during the mornings or afternoons of 23, 24 and 26 February, and 2 and 6 March 2015.

There were similar numbers of women (56.0%) and men (44.0%). Their mean age was 50 (range, 18-87) years. The majority (75.3%) had completed a secondary or higher level of education. Over half (53.3%) were non-religious, and approximately 40% were retired. The majority of men (63.6%, 42/66) and women (71.4%, 60/84) were married. Almost all participants (98.0%, n=147) declared themselves as heterosexual (Table 1).

In addition to our definitions of IPE, significantly more women than men also found chest examination of the respiratory system to be intimate (44.0%, 37/84 women vs 13.6%, 9/66 men; P<0.01). Among the entire sample, upper limb, lower limb, and abdominal examinations were also considered to be intimate by five (3.3%), 14 (9.3%), and 19 (12.7%) participants, respectively. Overall, 42 (50.0%) women preferred a female doctor for IPE; only one (1.2%) preferred a male doctor (Table 2). The only significant determining factor for women preferring a female doctor for IPE was the absence of prior experience of IPE (P=0.04); notable but non-significant determining factors included younger age (P=0.09) and being unmarried (P=0.10). For men, 42 (63.6%) participants did not have any preference for physician gender and 21 (31.8%) would prefer a male doctor (Table 2). No significant determining factors for men preferring either a male or female doctor for IPE were identified.

Of the 150 participants, 115 (76.7%) reported previous experience of IPE: 42 were men (63.6% of male participants) and 73 were women (86.9% of female participants). A large majority (90.7%, 136/150) of the participants considered the presence of a chaperone appropriate during IPE, and 84.0% (126/150) felt that doctors, irrespective of gender, should always ask if the patient would like a chaperone. This contrasted with the experience of those who had had previous IPE of whom only 72.6% (53/73) of women and 35.7% (15/42) of men reported having been chaperoned. For those whose previous IPE was unchaperoned, 75.0% (15/20) of women and 44.4% (12/27) of men would actually prefer to be chaperoned. Interestingly, participants with prior experience of IPE, irrespective of gender, were significantly more likely to want a chaperone when examined by a male doctor (57.4% with vs 28.6% without; P<0.01) but not a female doctor (28.7% with vs 25.7% without; P=0.87) when compared with those with no such experience. Most participants (72.7%, 109/150) regarded health care workers as suitable chaperones, and 52.7% (79/150) would also consider a family member, and 20.0% (30/150) a friend to be appropriate.

The most commonly cited reasons for preferring a chaperone included the availability of an objective account should any legal issue arise (61.3%, 92/150), protection against sexual harassment by the doctor (48.0%, 72/150), and psychological support (43.3%, 65/150) [Fig a]. The most commonly cited reasons for not having a chaperone included embarrassment (34.7%, 52/150); significantly more men (43.9%, 29/66) than women (27.4%, 23/84) considered a chaperone’s presence embarrassing (P=0.03). Furthermore, 26.0% of the participants (39/150) felt that the presence of a chaperone would undermine their privacy (Fig b).

The majority of women (75.0%, 63/84) would prefer a chaperone to be present when being examined by a male doctor, and 79.7% (55/69) of female respondents preferred that chaperone to be female (Table 2). Even when being examined by a female doctor, 28.6% (24/84) would still prefer to be chaperoned (Table 2). Women aged between 40 and 59 years were significantly more likely than other age-groups to prefer chaperoned IPE when examined by a female doctor (P=0.04). Other demographic factors—including education level, income, religion and marital status—were not significantly associated with any particular preferences.

More than half of the men had no specific preference, regardless of physician gender. There was, however, a proportion who would want to be chaperoned when examined by a male (21.2%, 14/66) or female (25.8%, 17/66) doctor (Table 2). Men who wished to be chaperoned when examined by a male doctor were significantly more likely to also prefer being chaperoned when examined by a female doctor (P<0.01). Of note is that a great majority of the men who preferred chaperoned IPE when examined by a male doctor indicated that chaperones could provide an objective account should a legal issue arise (85.7%, 12/14). There were no significant differences in terms of demographic variables between the ‘no preference’, ‘no chaperone’, and ‘prefer chaperone’ groups, regardless of the examining doctor’s gender.

Our results highlighted the different views of male and female subjects in this locality. The majority of our female patients would prefer to be chaperoned when examined by a male doctor, and previous experience of IPE appeared to enforce such a tendency. This finding is consistent with those from other countries.4 8 12 Interestingly, and somewhat unexpectedly, a substantial minority of women would still prefer to be chaperoned when examined by a female doctor. Another noteworthy finding was that male participants who preferred a chaperoned IPE by a male doctor would also hold a similar preference when examined by a female doctor, and that their most commonly cited reason was the availability of an objective account in case of medicolegal disputes. This was despite concerns about personal privacy and feelings of embarrassment.

Our findings compare well with the situation in the United Kingdom13 and Australia14 where the majority of patients were aware of and would prefer the practice. When compared with a similar study conducted at the Emergency Department in the US, however, we found that patients in Hong Kong were more likely to prefer chaperoned IPE when examined by a doctor of the opposite sex (male patient–female doctor: 25.8% vs 2-3%; female patient–male doctor: 75% vs 45-47%).3 Other studies found that physician gender had a variable impact on female patient’s preference for a chaperone.10 Our findings suggest a generally higher degree of acceptance of, if not expectation for, the presence of a medical chaperone in Hong Kong, although differences in study setting, subject profile, and study definitions of IPE limit the validity of such comparisons.

The reported experiences of participants who have had previous IPE indicate that their expectations have not been met on a number of occasions. Several reasons have been suggested for the inadequate use of chaperones including, inter alia, the shortage of nursing staff to act as chaperones,15 and a general lack of awareness.16 This is certainly an area for improvement in our health care setting. In this regard, it is important to note that some patients would also consider examinations of the abdomen and limbs to be intimate, and that chaperones may still be preferred even when the physician and patient are of the same gender. The Code of Professional Conduct issued by the Medical Council of Hong Kong states that ‘an intimate examination of a patient is recommended to be conducted in the presence of a chaperone to the knowledge of the patient’.17 ‘Intimate examination’ is not defined and there is no advice about the impact of gender. The United Kingdom General Medical Council (GMC) guidelines nonetheless specify that IPE may include ‘any examination where it is necessary to touch or even be close to the patient’, and the requirement for a chaperone should apply whether or not the patient and doctor are of the same or opposite gender.1 Challenging situations may arise when there is a shortage of staff or when the patient refuses the presence of a chaperone. The GMC have provided some detailed guidance.1

Our study has several limitations. First, the number of participants was relatively small and our findings may not be readily generalisable, particularly outside the public hospital setting, where patients are likely to be more familiar with their doctor and have more control over which doctor sees and examines them. The convenient sampling method that we adopted may result in systematic bias, skewed results, and potentially suboptimal generalisability of our findings. Second, our definition of IPE encompassed a range of different examinations, and it might be possible that participants have different preferences regarding each type. Third, our cohort consisted of relatively few young subjects (only 15% were aged <30 years) and this might have affected our ability to demonstrate any impact of age. Last, as to the reasons for participants’ preferences, our questionnaire only provided a short list of options rather than an open question and this could have limited the range of responses. Future studies may focus on patient’s preferences in specific settings (eg primary care) as well as physician’s practice in order to inform and promote public and professional awareness in Hong Kong.

Patients in Hong Kong have a high degree of acceptance towards the presence of a medical chaperone. Both female and male patients prefer such practice regardless of physician gender although individual patients may value the practice differently. Doctors are strongly encouraged to discuss the issue openly with their patients and offer the presence of a chaperone prior to any IPE; an alternative would be to put up a sign asking patients to notify the doctor or other staff if they prefer the presence of a chaperone during IPE.

All authors have disclosed no conflicts of interest.

1. General Medical Council. Good medical practice (2013). Available from: http://www.gmc-uk.org/guidance/good_medical_practice.asp. Accessed 27 Nov 2015.

2. Wai D, Katsaris M, Singhal R. Chaperones: are we protecting patients? Br J Gen Pract 2008;58:54-7. Crossref

3. Buchta RM. Adolescent females’ preferences regarding use of a chaperone during a pelvic examination. Observations from a private-practice setting. J Adolesc Health Care 1986;7:409-11. Crossref

4. Patton DD, Bodtke S, Homer RD. Patient perceptions of the need for chaperones during pelvic exams. Fam Med 1990;22:215-8.

5. Fiddes P, Scott A, Fletcher J, Glasier A. Attitudes towards pelvic examination and chaperones: a questionnaire survey of patients and providers. Contraception 2003;67:313-7. Crossref

6. Simanjuntak C, Cummings R, Chen MY, Williams H, Snow A, Fairley CK. What female patients feel about the offer of a chaperone by a male sexual health practitioner. Int J STD AIDS 2009;20:165-7. Crossref

7. Sinha S, De A, Jones N, Jones M, Williams RJ, Vaughan-Williams E. Patients’ attitude towards the use of a chaperone in breast examination. Ann R Coll Surg Engl 2009;91:46-9. Crossref

8. Penn MA, Bourguet CC. Patients’ attitudes regarding chaperones during physical examinations. J Fam Pract 1992;35:639-43.

9. Osmond MK, Copas AJ, Newey C, Edwards SG, Jungmann E, Mercey D. The use of chaperones for intimate examinations: the patient perspective based on an anonymous questionnaire. Int J STD AIDS 2007;18:667-71. Crossref

10. Santen SA, Seth N, Hemphill RR, Wrenn KD. Chaperones for rectal and genital examinations in the emergency department: what do patients and physicians want? South Med J 2008;101:24-8. Crossref

11. Van Hecke O, Jones K. The use of chaperones in general practice: Is this just a ‘Western’ concept? Med Sci Law 2015;55:278-83. Crossref

12. Teague R, Newton D, Fairley CK, et al. The differing views of male and female patients toward chaperones for genital examinations in a sexual health setting. Sex Transm Dis 2007;34:1004-7. Crossref

13. Pydah KL, Howard J. The awareness and use of chaperones by patients in an English general practice. J Med Ethics 2010;36:512-3. Crossref

14. Baber JA, Davies SC, Dayan LS. An extra pair of eyes: do patients want a chaperone when having an anogenital examination? Sex Health 2007;4:89-93. Crossref

15. Price DH, Tracy CS, Upshur RE. Chaperone use during intimate examinations in primary care: postal survey of family physicians. BMC Fam Pract 2005;6:52. Crossref

16. Vogel L. Chaperones: friend or foe, and to whom? CMAJ 2012;184:642-3. Crossref

17. Medical Council of Hong Kong. Code of Professional Conduct. 2009. Available from: http://www.mchk.org.hk/Code_of_Professional_Conduct_2009.pdf. Accessed 30 Aug 2015.

Ductal carcinoma in situ (DCIS) of the breast is a non-invasive, pre-cancerous lesion that was uncommon prior to the widespread use of mammography (MMG) screening; it is traditionally treated by mastectomy (MTX) with cure rates approaching 100%.1 The high incidence rate and mortality rate of breast cancer in women2 has led to the setting up of population-based breast cancer screening programmes by government in 34 countries.3 4 5 6 One of the results of the popularity of breast screening is the rise in the detected incidence of DCIS.7 Some western studies revealed that DCIS constituted approximately 10% to 40% of lesions detected by MMG screening.8 In Asia, following the pilot Singapore Breast Screening Project, the diagnosis of DCIS also increased markedly.9 Screen-detected DCIS showed a better clinical profile such as smaller size and higher chance of being treated by breast-conserving surgery (BCS).9 Early detection of breast cancer at this stage offers the best opportunity for curative treatment.10

The treatment of DCIS has been widely studied and reported in the western and other Asian countries.11 12 13 14 15 16 Although there have been no prospective randomised trials to compare MTX with BCS for DCIS, BCS has been widely accepted as an alternative treatment,17 especially for small mammographically detected lesions. In Hong Kong, however, data on DCIS are relatively limited. The Hong Kong Cancer Registry has only started to release basic data of annual incidence and age distribution of DCIS since 2009. In 2012, 3508 women in Hong Kong were diagnosed with invasive breast cancer and 477 women were diagnosed with DCIS18 that constituted only 12.0% of the total number of breast cancer patients diagnosed. This incidence of DCIS was relatively low compared with the 20.7% in the United States.14 Since our presentation and treatment pattern of DCIS are likely different to that in other countries, it is necessary to examine the particular pattern of care of DCIS in Hong Kong to better understand this disease.

The aim of this study was to evaluate the latest detection and treatment pattern for DCIS in Hong Kong so as to guide planning of future service provision.

The Hong Kong Breast Cancer Registry (HKBCR) was first established in 2007 by the Hong Kong Breast Cancer Foundation as a data collection and monitoring system for breast cancer in Hong Kong. The HKBCR aims to collect and analyse data from all Hong Kong breast cancer patients to obtain comprehensive information about demographics, risk exposures, treatments, clinical outcomes, and psychosocial impact on patients. It is the first population-wide, cancer-specific registry for breast cancer patients in Hong Kong and has been a member of the International Association of Cancer Registries since 2011, providing international standard data management and accuracy.

Between 2008 and 2011, a total of 5393 patients with a history of in-situ or invasive breast cancers were registered with the HKBCR on a voluntary basis. Of these patients, 2539 (47.1%) and 2854 (52.9%) were recruited from private clinics or hospitals and public hospitals, respectively. Demographics and risk exposure data were collected from these patients by questionnaire; clinical characteristics, detection methods, diagnostic methods, disease stage, histopathological profile, treatment modalities, and clinical outcome data were extracted from their medical records.19 Data analysis was carried out in December 2014.

Inclusion criteria for this study were as follows: female patient being diagnosed and treated in Hong Kong from January 2001 to December 2011; pure DCIS with no invasive element in ipsilateral or contralateral breast at the time of diagnosis; definitive surgery performed; complete pathology details available; if axillary node sampling/dissection was performed, the nodal status must be negative; and no prior neoadjuvant treatment administered. Overall, 573 patients, including 16 synchronous patients with bilateral DCIS, from the HKBCR fulfilled the above criteria for further study.

For comparison purposes, the records of female patients with invasive breast cancer diagnosed and treated in Hong Kong during the same period were also extracted from HKBCR. Altogether, 1611 invasive breast carcinoma patients with 20 synchronous bilateral patients were retrieved for data analysis.

In this study, local recurrence was defined as the reappearance of cancer, invasive or non-invasive, in the treated breast or chest wall before or at the time of regional or distant metastases. All events were measured from the date of the definitive surgery. Descriptive statistics were used to describe the patterns of demographic and pathological features. Statistical significance was tested using Chi squared tests for categorical variables. The Kaplan-Meier method was applied to analyse the local recurrence estimation. All statistical tests were two-sided and performed at the 0.05 level of significance (P value). The Statistical Package for the Social Sciences (Windows version 19.0; SPSS Inc, Chicago [IL], US) was used for all statistical analyses.

The project was approved by respective Institutional Review Board and Ethics Committee of the following hospitals: Hong Kong Sanatorium & Hospital, Hong Kong Baptist Hospital, Hong Kong Adventist Hospital, Princess Margaret Hospital, United Christian Hospital, Prince of Wales Hospital, Queen Mary Hospital, Pamela Youde Nethersole Eastern Hospital, Pok Oi Hospital, North District Hospital, Tuen Mun Hospital, and Yan Chai Hospital. All participants provided written informed consent.

Of the 573 patients with DCIS of breast, the majority (74.9%) were diagnosed and treated between 2006 and 2011. A similar distribution was found in the 1611 patients with invasive breast cancer.

Table 1 compares the demographic characteristics of DCIS and invasive breast cancer patients. The results indicate that DCIS patients were significantly younger (median, 48.6 vs 50.3 years; P<0.001), had a higher education level (matriculation or above, 34.0% vs 15.0%; P<0.001), were more likely to be working (61.1% vs 55.8%; P<0.001), and had a higher total monthly family income of HK$30 000 or above (32.6% vs 16.6%; P<0.001). More DCIS patients had regular breast screening habits in the form of self-examination (monthly: 23.6% vs 21.8%; P<0.001), clinical breast examination (yearly: 41.5% vs 27.3%; P<0.001), MMG screening (yearly: 24.1% vs 10.6%; P<0.001), and breast ultrasound screening (yearly: 20.9% vs 10.2%; P<0.001). Patients with DCIS had a much higher chance of being asymptomatic at diagnosis (ie screen-detected) than their invasive breast cancer counterparts (29.0% vs 4.7%; P<0.001). Significantly more DCIS patients underwent BCS than their invasive cancer counterparts (55.8% vs 36.7%; P<0.001). Among those treated by BCS, DCIS patients had a similar chance of receiving adjuvant radiotherapy as the invasive cancer patients (94.1% vs 93.8%). As expected, only very few (8.7%) DCIS patients required adjuvant radiotherapy after MTX. Although DCIS patients do not require systemic adjuvant therapy, some may be prescribed hormone therapy as chemoprevention. In our study, only a small percentage of DCIS patients received hormone therapy and the pattern was similar after BCS or MTX (19.4% after BCS and 17.0% after MTX; P<0.001).

Table 2 shows the patient demographics, and clinical and pathological characteristics of screen-detected (asymptomatic) and self-detected (symptomatic) DCIS in Hong Kong. There was no significant difference in the median age between these subgroups (49.1 vs 48.5 years; P=0.23). The screen-detected subgroup had a significantly higher education level (matriculation or above, 42.7% vs 29.6%; P=0.01), higher total monthly family income of HK$30 000 or above (45.7% vs 26.8%; P=0.01), and underwent more regular clinical breast examination (yearly: 49.4% vs 35.5%; P<0.001), MMG (every 2 years: 22.0% vs 7.0%; P<0.001), and breast ultrasound screening (every 2 years: 11.6% vs 3.4%; P<0.001).

Among the DCIS patients, 28.6% (164/573) were screen-detected: since MMG screening is not usually recommended in younger women, only 14.5% (11/76) of DCIS in patients aged below 40 years were screen-detected compared with 34.3% (148/431) in patients aged above 40 years. Irrespective of the type of presentation, two thirds or more of DCIS patients chose to have surgery at a private hospital and the screen-detected subgroup had an even higher tendency to do so (74.4% vs 65.6%; P=0.05). Although there was a trend of finding smaller lesions in the screen-detected subgroup, there was no significant difference between the two subgroups in tumour size (median: 1.6 cm vs 2.0 cm; P=0.14). Despite this finding, screen-detected DCIS patients had a higher chance of being treated by BCS than symptomatic patients (65.9% vs 51.4%; P=0.02) [Table 2].

Among 573 patients with DCIS, clinical outcome data were available for 487 patients only. With a median follow-up of 3.1 (range, 0.5-10.9) years, the early clinical outcome was very good and compatible with other series. The overall crude local recurrence rate in DCIS patients was 3.9% (19 in 487 patients) and, as expected, there was a significant difference between MTX and BCS patients (0.4% vs 3.3%) [Table 3]. Of the 18 patients with known pathology at recurrence, eight (44.4%) had developed invasive components. Of the 16 BCS patients, 11 (68.8%) underwent salvage MTX at recurrence. Overall, by 6 years, the projected local recurrence rates after BCS were similar for DCIS patients and invasive breast cancer patients (log rank, P=0.21; Fig). No regional recurrence, distant recurrence, or cancer-related death were observed in the DCIS patients.

Ductal carcinoma in situ was relatively uncommon in western countries until the widespread use of mass breast screening. There is strong evidence that the recent rise in DCIS incidence is related to the popularity of breast screening. Since there is no government-funded breast screening programme in Hong Kong, our study showed that only 29.0% of the DCIS in Hong Kong was first detected by screening, significantly lower than the 80% screen-detected rate in DCIS of other studies.20 21 22 23 24 Our data showed that these DCIS patients in general had a higher monthly family income and higher level of education than their invasive cancer counterparts and this may contribute to their higher acceptance of self-funded breast screening, higher breast cancer awareness, and hence better chance of detecting breast cancer at an early stage. Not surprisingly, the use of BCS was also significantly more popular in the DCIS patients compared with their invasive cancer counterparts (55.8% vs 36.7%; P<0.001). Our results were also consistent with a previous local study that reported the performance of opportunistic breast screening in local well women clinics and showed that breast screening could achieve a higher cancer detection rate and detect the cancer at an early stage.25

In contrast, the profile of screen-detected and self-detected (ie symptomatic) DCIS patients showed more similarities than differences. The overall tumour size was not significantly different between these subgroups, although lesions of >2 cm were less common in the screen-detected patients (34.8% vs 43.0%; P=0.12). Since there is no population-based breast screening programme in Hong Kong, these opportunistic breast screenings performed in various laboratories may have inherent limitations. Overseas studies have shown a considerably higher sensitivity in organised population-based screening than in opportunistic screening,24 although a large-scale local self-referred breast screening centre reported comparable performance.25

Our study showed a high acceptance of BCS for management of DCIS in Hong Kong and nearly all (94.1%) BCS patients also underwent postoperative radiotherapy. Although prior randomised studies have demonstrated the benefit of postoperative radiotherapy in reducing both invasive and non-invasive recurrence of DCIS after BCS, much effort has been put into identifying a low-risk subgroup in whom postoperative radiotherapy can be safely omitted.26 The Van Nuys prognostic index (VNPI)—a retrospectively derived risk classification that combines tumour size, margin width, and pathological classification—was developed to select this low-risk group.27 Nonetheless, perhaps due to contradictory findings from other studies that reported a much higher local failure rate in the VNPI low-risk subgroup,28 it is apparent that most clinicians in Hong Kong had reservations when applying the VNPI to their DCIS patients.

Although tamoxifen after local excision for DCIS has been shown to reduce the risk of recurrent DCIS in the ipsilateral breast (hazard ratio=0.75; 95% confidence interval [CI], 0.61-0.92) and contralateral breast (relative risk=0.50; 95% CI, 0.28-0.87)29 and over 60% of our patients had positive hormonal receptors, less than 20% DCIS patients in Hong Kong actually received tamoxifen as chemoprevention.30 It is likely related to the concern about side-effects (particularly the small risk of endometrial cancer) and the lack of overall survival benefit as shown by the Cochrane systematic review and meta-analysis.29 The lack of survival benefit is consistent with the clinical experience that most new lesions detected during follow-up surveillance are highly treatable.

As expected, the local recurrence rate after MTX was very low (0.4%) in these DCIS patients; it should be noted that 8.7% had received adjuvant radiotherapy, probably because of close resection margins. For DCIS patients treated by BCS, the crude local recurrence rate in our study was 3.3% (16 in 320 patients) and was quite similar to the long-term experience in another regional hospital (Pamela Youde Nethersole Eastern Hospital) in Hong Kong. In their analysis of 155 DCIS patients treated by BCS and radiotherapy, after a 10-year median follow-up, the crude local recurrence rate was 5.8% (unpublished data of Pamela Youde Nethersole Eastern Hospital). Our study did not capture the data on the mode of detection of local recurrences but another local study reported that only 43% of in-breast recurrences could be first detected by surveillance breast imaging; the rest presented with either nipple discharge or a palpable mass.30 Hence, patients should be advised not to become overly dependent on breast imaging to detect early recurrences. Although there were no cancer-related deaths in these DCIS patients, 44% of local recurrences in this study contained invasive components that may still necessitate systemic treatment in addition to further salvage surgery.

This study provides the first comprehensive analysis of the pattern of care of DCIS in Hong Kong. The strength of this analysis is the comprehensiveness compared with other cancer registries in data collection on epidemiological, pathological, and treatment characteristics for breast cancer. Nonetheless, data from HKBCR might not be representative of all breast cancers in Hong Kong since a higher proportion of patients were recruited from private hospitals or clinics than public hospitals in Hong Kong. Since the data collection was done on a voluntary basis and only started in 2008, some clinical outcome data may be missing (approximately 10% of DCIS patients) and the follow-up duration remains relatively short, and may not represent the whole local population. There was also a high proportion of missing data for family income in the two internal comparisons. Although the distribution of age at diagnosis in our study did not deviate too far from that reported in the Hong Kong Cancer Registry (a population-based registry; Table 4), older age-groups, especially those aged 70 years and above, were under-represented in the present study. Furthermore, we did not have information on education, occupation, and family income to enable comparison of socio-economic backgrounds.

Ductal carcinoma in situ in Hong Kong appears to be a more prevalent disease in the higher social classes with a tendency to be managed in the private sector. More than half of DCIS patients can be successfully treated with BCS and the early outcome is excellent and comparable with overseas studies.9 16 Further studies are needed to examine the long-term clinical outcome of DCIS in Hong Kong.

The authors thank all of the patients/survivors, doctors, and research staff who have participated in HKBCR to facilitate data collection at clinics and hospitals throughout the territory. The authors also acknowledge the following steering committee members: Prof Emily Chan, Dr John Chan, Dr Keeng-Wai Chan, Dr Miranda Chan, Dr Sharon Chan, Dr Foon-Yiu Cheung, Dr Peter Choi, Dr Josette Chor, Ms Yvonne Chua, Ms Doris Kwan, Dr Wing-Hong Kwan, Dr Stephen Law, Dr Simon Leung, Dr Lawrence Li, Dr Janice Tsang, Dr Gary Tse, Mrs Cecilia Tseung, Dr Tung Yuk, Ms Lorna Wong, Dr Ting-Ting Wong, Dr CC Yau, Prof Winnie Yeo, and Prof Benny Zee for providing guidance for the development of the HKBCR.

All authors have disclosed no conflicts of interest.

1. Fortunato L, Poccia I, de Paula U, Santini E. Ductal carcinoma in situ: what can we learn from clinical trials? Int J Surg Oncol 2012;2012:296829.

2. International Agency for Research on Cancer. Globocan 2012: All cancers (excluding non-melanoma skin cancer)—estimated incidence, mortality and prevalence worldwide in 2012. Available from: http://globocan.iarc.fr/Pages/fact_sheets_cancer.aspx. Accessed 2 Apr 2015.

3. Kwong A, Cheung PS, Wong AY, et al. The acceptance and feasibility of breast cancer screening in the East. Breast 2008;17:42-50. Crossref

4. Biesheuvel C, Weigel S, Heindel W. Mammography screening: evidence, history and current practice in Germany and other European countries. Breast Care (Basel) 2001;6:104-9. Crossref

5. Cheung PS. Global and local status of breast cancer and screening. Paper presented at Hong Kong Breast Cancer Foundation Seminar; 2013 Apr 14; Hong Kong.

6. Hellquist BN, Duffy SW, Abdsaleh S, et al. Effectiveness of population-based service screening with mammography for women ages 40 to 49 years: evaluation of the Swedish Mammography Screening in Young Women (SCRY) cohort. Cancer 2011;117:714-22. Crossref

7. Leonard GD, Swain SM. Ductal carcinoma in situ, complexities and challenges. J Natl Cancer Inst 2004;96:906-20. Crossref

8. Delaney G, Ung O, Bilous M, Cahill S, Greenberg M, Boyages J. Ductal carcinoma in situ. Part I: definition and diagnosis. Aust N Z J Surg 1997;67:81-93. Crossref

9. Chuwa EW, Yeo AW, Koong HN, et al. Early detection of breast cancer through population-based mammographic screening in Asian women: a comparison study between screen-detected and symptomatic breast cancers. Breast J 2009;15:133-9. Crossref

10. Morrow M, Strom EA, Bassett LW, et al. Standard for the management of ductal carcinoma in situ of the breast (DCIS). CA Cancer J Clin 2002;52:256-76. Crossref

11. Irvine T, Fentiman IS. Biology and treatment of ductal carcinoma in situ. Expert Rev Anticancer Ther 2007;7:135-45. Crossref

12. Siziopikou KP. Ductal carcinoma in situ of the breast: current concepts and future directions. Arch Pathol Lab Med 2013;137:462-6. Crossref

13. Jiveliouk I, Corn B, Inbar M, Merimsky O. Ductal carcinoma in situ of the breast in Israeli women treated by breast-conserving surgery followed by radiation therapy. Oncology 2009;76:30-5. Crossref

14. Virnig BA, Torchia MT, Jarosek SL, Durham S, Tuttle TM. Data Points #14: Use of endocrine therapy following diagnosis of ductal carcinoma in situ or early invasive breast cancer. Rockville (MD): Agency for Healthcare Research and Quality; 2011-2012.

15. Tan KB, Lee HY, Putti TC. Ductal carcinoma in situ of the breast in Singapore: recent trends and clinical implications. ANZ J Surg 2002;72:793-7. Crossref

16. Chuwa EW, Tan VH, Tan PH, Yong WS, Ho GH, Wong CY. Treatment for ductal carcinoma in situ in an Asian population: outcome and prognostic factors. ANZ J Surg 2008;78:42-8. Crossref

17. Mouw KW, Harris JR. Irradiation in early-stage breast cancer: conventional whole-breast, accelerated partial-breast, and accelerated whole-breast strategies compared. Oncology (Williston Park) 2012;26:820-30.

18. Hong Kong Cancer Registry. Leading cancer sites in Hong Kong in 2012. Available from: http://www3.ha.org.hk/cancereg/rank_2012.pdf. Accessed 15 Jan 2015.

19. Cheung P, Hung WK, Cheung C, et al. Early data from the first population-wide breast cancer-specific registry in Hong Kong. World J Surg 2012;36:723-9. Crossref

20. Chua MS, Mok TS, Kwan WH, Yeo W, Zee B. Knowledge, perceptions, and attitudes of Hong Kong Chinese women on screening mammography and early breast cancer management. Breast J 2005;11:52-6. Crossref

21. Schouten van der Velden AP, Peeters PH, Koot VC, Hennipman A. Clinical presentation and surgical quality in treatment of ductal carcinoma in situ of the breast. Acta Oncol 2006;45:544-9. Crossref

22. Solin LJ, Fourquet A, Vicini FA, et al. Long-term outcome after breast-conservation treatment with radiation for mammographically detected ductal carcinoma in situ of the breast. Cancer 2005;103:1137-46. Crossref

23. Habermann EB, Abbot A, Parsons HM, Virnig BA, Al-Refaie WB, Tuttle TM. Are mastectomy rates really increasing in the United States? J Clin Oncol 2010;28:3437-41. Crossref

24. Bihrmann K, Jensen A, Olsen AH, et al. Performance of systematic and non-systematic (‘opportunistic’) screening mammography: a comparative study from Denmark. J Med Screen 2008;15:23-6. Crossref

25. Lui CY, Lam HS, Chan LK, et al. Opportunistic breast cancer screening in Hong Kong; a revisit of the Kwong Wah Hospital experience. Hong Kong Med J 2007;13:106-13.

26. Fisher B, Dignam J, Wolmark N, et al. Lumpectomy and radiation therapy for the treatment of intraductal breast cancer: findings from National Surgical Adjuvant Breast and Bowel Project B-17. J Clin Oncol 1998;16:441-52.

27. Silverstein MJ, Lagios MD, Craig PH, et al. A prognostic index for ductal carcinoma in situ of the breast. Cancer 1996;77:2267-74. Crossref

28. de Mascarel I, Bonichon F, MacGrogan G, et al. Application of the Van Nuys prognostic index in a retrospective series of 367 ductal carcinomas in situ of the breast examined by serial macroscopic sectioning: practical considerations. Breast Cancer Res Treat 2000;61:151-9. Crossref

29. Staley H, McCallum I, Bruce J. Postoperative tamoxifen for ductal carcinoma in situ: Cochrane systematic review and meta-analysis. Breast 2014;23:546-51. Crossref

30. Yau TK, Sze H, Soong IS, et al. Surveillance mammography after breast conservation therapy in Hong Kong: effectiveness and feasibility of risk-adapted approach. Breast 2008;17:132-7. Crossref

Prostate cancer is the second most frequently diagnosed cancer of men worldwide, with the highest incidence and prevalence rates occurring in more developed societies.1 The incidence of prostate cancer is also increasing in Asian countries.2 Many reasons have contributed to this recent rise in incidence in Asia, including the increase in the ageing population, the westernised diet, and also the increased use of serum prostate-specific antigen (PSA) for cancer detection.3 4 Although current evidence supports the use of PSA testing to decrease the incidence of metastatic disease and prostate cancer–specific mortality,5 the use of serum PSA for the early detection of prostate cancer is still controversial.6 7 One of the concerns is the risk of overdiagnosis and overtreatment of low-risk cancer that may result in more potential harm than benefit to patients.8 9 10 There are many types of prostate cancer screening approaches. Currently, there is no structured prostate cancer screening programme in Asia. Therefore, early diagnosis of prostate cancer in Asia is by an opportunistic case-finding approach, that is, offering PSA testing to an individual without obvious signs and symptoms of prostate cancer. Information on the characteristics of prostate cancer diagnosed by various approaches in Asia is lacking, however. We postulated that patients diagnosed by a case-finding approach, such as during a routine health check or a consultation for symptoms with a low suspicion of prostate cancer origin, would have a better prognosis than those who presented with symptoms related to prostate cancer, with or without metastatic disease. We investigated the relationship between the mode of presentation and the characteristics of prostate cancers diagnosed in our hospital.

This was a prospective cohort study to assess consecutive adult male patients diagnosed with prostate cancer at Prince of Wales Hospital, a regional hospital in Hong Kong, between September 2011 and February 2013. Institutional ethics approval was obtained for the study. Informed consent was obtained from all study subjects prior to enrolment in the study.

All patients aged 18 years or above at our hospital with histological confirmation of prostate cancer were identified and approached for inclusion in this study. After informed consent was obtained, information on the initial presentation of the patient’s condition, prostate cancer characteristics at diagnosis, and the initial treatment plan were collected. Patients were then followed up for a minimum of 2 years, and the clinical outcome was assessed. All cancer was graded using the Gleason grading system that is based on the histological pattern of the cancer tissue. The tissue was graded from 1 (well-differentiated) to 5 (poorly differentiated). Each biopsy was given two scores, the first indicated the most common pattern and the second, the highest grading.11 Our scoring system for prostate cancer consists of staging according to TNM staging 201012 and risk stratification according to the National Comprehensive Cancer Network (NCCN) guideline.13

Subjects were divided into four groups according to the initial clinical presentation of their prostate cancer by two investigators who were blinded to the clinical outcome during the assessment and then confirmed by a senior investigator. Any discrepancy was discussed and a final allocation made. The health check group (group 1) included patients in whom a raised PSA was detected during a routine health check. Group 2 comprised patients diagnosed with prostate cancer by the case-finding approach after they presented with symptoms with no/low clinical suspicion of prostate cancer (eg renal cysts, non-specific abdominal pain). Group 3 comprised patients with a high clinical suspicion of prostate cancer or malignant disease, for example, lower urinary tract symptoms with abnormal digital rectal examination (DRE), bone pain, or weight loss. Finally, those patients with a histological diagnosis of prostate cancer made following transurethral resection of the prostate (TURP) but with no preoperative suspicion of prostate cancer were assigned to the TURP group (group 4).

Since prostate cancer arises mostly from the peripheral zone (in contrast to benign prostate hyperplasia [BPH] that commonly arises from the transition zone), patients with early-stage prostate cancer are usually asymptomatic.14 Not until the tumour becomes locally advanced (with clinical signs of abnormal DRE) do patients have voiding symptoms attributed to prostate cancer. Therefore, in a patient who presents with lower urinary tract symptoms (LUTS) and normal DRE, the symptoms are more likely related to BPH, not secondary to prostate cancer. Testing of PSA is not routine for male patients with LUTS.15 16 According to the Guidelines from the European Association of Urology, PSA measurement should only be performed to assess the risk of progression of LUTS or if a diagnosis of prostate cancer would change disease management.16 For patients who present with LUTS but with a low clinical suspicion of prostate cancer (ie normal DRE), PSA testing is considered case-finding for prostate cancer. In this study, such patients were assigned to group 2. This also applied to other presenting symptoms with no or low suspicion of being related to prostate cancer. Nonetheless, in subjects with LUTS and clinical symptoms or signs suspected to be secondary to prostate cancer, such as abnormal DRE findings, PSA testing would be part of the diagnostic process for prostate cancer, not case-finding. As a result, these patients would be assigned to group 3.

After all data were collected, descriptive statistics were applied. A Chi squared test or Fisher’s exact test was used to determine any relationship between the categorical outcome measures. Analysis of variance or Kruskal-Wallis test was used for normal or skewed data, and then followed by post-hoc comparisons with Bonferroni adjustment. Kaplan-Meier survival analysis was applied to analyse survival among the four groups. Data management and analysis were performed using the Statistical Package for the Social Sciences (Windows version 22.0; SPSS Inc, Chicago [IL], US). A two-tailed test was used with significance set at P<0.05.

From September 2011 to February 2013, 126 consecutive patients with newly diagnosed, histologically confirmed prostate cancer were managed in our centre. One patient refused to participate in this study, and five patients were not capable of providing informed consent. Therefore 120 patients were enrolled in this study: group 1 (n=12), group 2 (n=53), group 3 (n=46), and group 4 (n=9) [Table 1]. The initial presenting symptoms of patients in groups 2 and 3 are listed in Table 2. In group 2, 43 patients presented with LUTS (including three with acute urinary retention) with low clinical suspicion of prostate cancer. Ten patients presented with other symptoms—seven with other urological symptoms and three with other general surgical problems. Among them, three patients (one with loin pain, one with erectile dysfunction, and one with hernia) were found to have abnormal DRE during consultation. In group 3, 33 patients presented with LUTS (nine patients with acute urinary retention) and abnormal DRE. Three patients presented with haematuria and DRE during initial workup was abnormal and a subsequent diagnosis was made of prostate cancer. Nine patients presented with metastatic symptoms, eg bone pain, acute spinal cord compression, and abnormal soft tissue mass. One patient presented with weight loss and was subsequently diagnosed to have non-metastatic prostate cancer (Table 2).

The overall mean age was 71.0 (range, 54-90) years (Table 1). Age and serum PSA level were statistically significantly different across groups. Multiple comparisons with Bonferroni correction revealed that patients in group 3 were significantly older than those in groups 1 and 2 (P<0.001). Patients in group 3 also had a significantly higher serum PSA level compared with those in group 1 (P=0.044) and group 2 (P=0.045) by Kruskal-Wallis test. In group 3, 41 (89.1%) patients had an abnormal DRE (P<0.001).

With regard to disease status, the numbers of patients with a Gleason sum of ≥7 were seven (58.3%) in group 1, 12 (22.6%) in group 2, 32 (69.6%) in group 3, and four (44.4%) in group 4. In accordance with the NCCN guideline, the number of patients with disease more severe than very low or low risk were eight (66.7%) in group 1, 30 (56.6%) in group 2, 43 (93.5%) in group 3, and seven (77.8%) in group 4 (Table 1). In group 3, 24 (52.2%) patients had metastatic disease at initial presentation, a much higher rate than in the other groups (P<0.001, Fisher’s exact test).

Since both groups 1 and 2 had no prostate cancer–related symptoms, we tried to combine the two groups to assess the characteristics of prostate cancer diagnosed by a case-finding approach. Group 3 patients had significantly older age, higher serum PSA level, more aggressive disease (Gleason sum ≥7), and more metastatic disease at presentation than the combined groups 1 and 2 patients (P<0.001 for all parameters; Table 3).

The types of primary treatment administered are listed in Table 4. The number of patients receiving radical local therapy (either surgery or radiotherapy) was 10 (83.3%) in group 1, 44 (83.0%) in group 2, 11 (23.9%) in group 3, and one (11.1%) in group 4. Systemic androgen deprivation therapy was prescribed to one (8.3%) patient in group 1, three (5.7%) in group 2, 26 (56.5%) in group 3, and two (22.2%) in group 4 (P<0.0005). Because group 3 had significantly more patients with locally advanced and metastatic disease, significantly fewer could be managed with curative-intent local therapy. Among those patients with very low– or low-risk disease in groups 1 and 2, one (25%) and five (21.7%) respectively elected to have conservative management, either watchful waiting or active surveillance.

The median follow-up period was 32 months (interquartile range, 28-35 months). No patients were lost to follow-up. Eleven (9.2%) patients died—10 (21.7%) in group 3 and one (11.1%) in group 4. No patient in groups 1 or 2 died during the follow-up period. The causes of death in group 3 patients were directly related to prostate cancer in seven patients, metastatic bladder cancer in one patient, and acute myocardial infarction in two patients. The cause of death of the patient in group 4 was secondary to advanced rectosigmoid carcinoma. Therefore, the overall rate of cancer-specific survival for the total cohort was 91.0%, but 100% for each of groups 1, 2, and 4 compared with only 76.8% for group 3 (P=0.006, log-rank test; Fig).

Since the introduction of PSA testing, there has been a worldwide change in the presentation of prostate cancer. More and more prostate cancers are diagnosed at a lower risk level and earlier stage for which curative treatment can be offered.17 18 As a result, the use of PSA testing for early detection of prostate cancer is believed to be one of the factors that has led to the decrease in overall prostate cancer mortality in many developed areas.2 From our cohort, we also observed that patients with prostate cancer diagnosed by case-finding approach using PSA testing (groups 1 and 2) had significantly more clinically localised disease and hence a higher chance of receiving curative-intent treatment than those patients who presented with prostate cancer–related symptoms (group 3).

We also observed that the short-term cancer-specific mortality rate of patients who presented with prostate cancer–related symptoms (group 3) was significantly higher than that in the other groups. In our cohort, more than half of the patients in group 3 already had metastasis at diagnosis. Because patients presenting with metastasis have a much poorer outcome than other patients,19 20 it was not surprising that the mortality rate for patients who presented with symptoms was also higher. This indirectly supports the case-finding approach by PSA testing in patients with symptoms but no/low clinical suspicion of prostate cancer, as it might help to decrease the incidence of metastatic disease and hence the mortality related to prostate cancer.21

Although PSA testing is widely performed in western countries to detect early prostate cancer, its use in Asian countries is still not a common practice. From a population-based telephone survey involving 1002 Chinese men aged ≥50 years in Hong Kong, only 9.5% had ever had a PSA test, and only 3.7% of the total sample had PSA test done during a routine health check.22 Even in more developed Asian countries such as Japan and South Korea, only 15% to 20% of the population had had a PSA test.23 Only 10% of the prostate cancers in our cohort were diagnosed during a self-initiated health check with PSA testing. Therefore, offering a PSA test as case-finding for prostate cancer during a patient’s consultation for non-prostate-cancer–related symptoms is an alternative approach for early detection of prostate cancer. We believe that this case-finding approach is feasible for the detection of early prostate cancer in our region, where public awareness and use of PSA testing is still low. Certainly, patients need to be well informed about the nature and implications of PSA testing before the test is performed.24 25

The main concerns surrounding the use of PSA testing for the detection of early prostate cancer are overdiagnosis and overtreatment.8 9 10 Only approximately 36% of patients in the study cohort had very low– or low-risk disease that might not require aggressive intervention.13 26 Even in those patients with prostate cancer diagnosed by PSA testing (ie groups 1 and 2), more than 50% were in the intermediate- or higher-risk groups. Testing of PSA level did help to detect patients with significantly high-risk prostate cancer that warranted further treatment. To minimise the risk of overtreatment, the Melbourne Consensus Statement advises uncoupling of the prostate cancer diagnosis from the intervention.5 Offering active surveillance to patients with low-risk disease will help to minimise the potential harm of overtreatment. In our cohort, for patients in groups 1 and 2 with very low– or low-risk disease, six (22.2%) opted for observation with no active treatment. We believe this concept should be promoted to both clinicians and patients, rather than limiting the use of PSA testing for the case finding of prostate cancer.

Currently, some newly proposed strategies, such as determination of the baseline PSA level earlier in life27 28 and the use of newer diagnostic tools,29 30 might help to reduce unnecessary prostatic biopsies and overdiagnosis of low-risk prostate cancer. Nonetheless, since most of these studies were conducted in Caucasian-based populations, further studies in Asian populations are necessary to verify their suitability in our region.

Although our data show that the short-term outcome of patients who present with prostate cancer–related symptoms seems to be worse than those diagnosed by PSA testing, this might be due to potential lead-time bias, that is, the increase in survival is actually due to the length of time between the detection of a disease by PSA testing and its usual clinical presentation and diagnosis. This will result in an increase in survival time for patients diagnosed by PSA testing. Other potential bias is length-time bias which suggests that annual PSA may only detect slow-growing tumours, that screening for prostate cancer does not detect the very tumour for which it is intended.

The aim of our study was not to assess the role of PSA testing in the detection of early prostate cancer or its effect on long-term outcome and survival of patients. Rather, we aimed only to compare cancer characteristics and short-term outcome among patients with different presentations. We also did not analyse the potential harm of PSA testing, prostatic biopsy, or morbidities related to treatment. The positive rate for prostatic biopsy depends on the level of serum PSA and DRE finding. From local experience, for patients with a normal DRE, the positive rate of prostatic biopsy for serum PSA level of 4-10 ng/mL and 10-20 ng/mL was only 6.7%-13.4% and 10.3%-21.8%, respectively.31 32 33 Therefore, information on this would be helpful during patient counselling for prostatic biopsy. Another limitation of our study was the relatively small sample size from a single centre in Hong Kong, therefore our results might not represent the general situation in Hong Kong. Further studies, especially with multicentre collaboration, may help to confirm the applicability of our results in the local population.

Patients with prostate cancer presenting with related symptoms had more metastatic disease and poorer survival than those diagnosed by a case-finding approach using PSA testing during a health check or management of symptoms with a low suspicion of prostate cancer. More than half of the patients diagnosed by this case-finding approach belonged to intermediate- or higher-risk groups for which active treatment was recommended. Apart from a self-initiated health check with PSA testing, offering PSA testing to appropriate patients who present with symptoms with no/low clinical suspicion of prostate cancer is an alternative approach to early diagnosis of prostate cancer. Pre-test counselling, including the discussion of potential bias (such as lead time or length-time bias), is essential. This may hopefully help to improve the short-term outcome for these patients.

All authors have disclosed no conflicts of interest.

1. Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer statistics, 2012. CA Cancer J Clin 2015;65:87-108. Crossref

2. Center MM, Jemal A, Lortet-Tieulent J, et al. International variation in prostate cancer incidence and mortality rates. Eur Urol 2012;61:1079-92. Crossref

3. Sim HG, Cheng CW. Changing demography of prostate cancer in Asia. Eur J Cancer 2005;41:834-45. Crossref

4. Pu YS, Chiang HS, Lin CC, Huang CY, Huang KH, Chen J. Changing trends of prostate cancer in Asia. Aging Male 2004;7:120-32. Crossref

5. Murphy DG, Ahlering T, Catalona WJ, et al. The Melbourne Consensus Statement on the early detection of prostate cancer. BJU Int 2014;113:186-8. Crossref

6. Schröder FH. Landmarks in prostate cancer screening. BJU Int 2012;110 Suppl 1:3-7. Crossref

7. Alberts AR, Schoots IG, Roobol MJ. Prostate-specific antigen-based prostate cancer screening: past and future. Int J Urol 2015;22:524-32. Crossref

8. Welch HG, Black WC. Overdiagnosis in cancer. J Natl Cancer Inst 2010;102:605-13. Crossref

9. Loeb S, Bjurlin MA, Nicholson J, et al. Overdiagnosis and overtreatment of prostate cancer. Eur Urol 2014;65:1046-55. Crossref

10. Stone NN, Crawford ED. To screen or not to screen: the prostate cancer dilemma. Asian J Androl 2015;17:44-5. Crossref

11. Epstein JI. An update of the Gleason grading system. J Urol 2010;183:433-40. Crossref

12. Edge S, Byrd DR, Compton CC, Fitz AG, Greene FL, Trotti A. AJCC cancer staging manual. 7th ed. New York, NY: Springer; 2010.

13. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology—Prostate cancer, version 1.2015. Available from: http://www.nccn.org. Accessed 1 Jul 2015.

14. Loeb S, Carter HB. Early detection, diagnosis, and staging of prostate cancer. In: Wein AJ, Kavoussi LR, Novick AC, et al, editors. Campbell-Walsh Urology. 10th ed. Philadelphia, PA: Elsevier Saunders; 2012: 2763-70. Crossref

15. European Association of Urology. Guidelines on the management of non-neurogenic male lower urinary tract symptoms (LUTS), incl. benign prostatic obstruction (BPO). Available from: http://uroweb.org/wp-content/uploads/EAU-Guidelines-Non-Neurogenic-Male-LUTS_LR.pdf. Accessed 1 Jul 2015.

16. Arianayagam M, Arianayagam R, Rashid P. Lower urinary tract symptoms—current management in older men. Aust Fam Physician 2011;40:758-67.

17. Cooperberg MR, Lubeck DP, Mehta SS, Carroll PR, CaPSURE. Time trends in clinical risk stratification for prostate cancer: implications for outcomes (Data from CaPSURE). J Urol 2003;170(6 Pt 2):S21-7. Crossref

18. Crawford ED. Epidemiology of prostate cancer. Urology 2003;62(6 Suppl 1):3-12. Crossref

19. Chowdhury S, Robinson D, Cahill D, Rodriguez-Vida A, Holmberg L, Møller H. Causes of death in men with prostate cancer: an analysis of 50,000 men from the Thames Cancer Registry. BJU Int 2013;112:182-9. Crossref

20. Patrikidou A, Loriot Y, Eymard JC, et al. Who dies from prostate cancer? Prostate Cancer Prostatic Dis 2014;17:348-52. Crossref

21. Wu JN, Fish KM, Evans CP, Devere White RW, Dall’Era MA. No improvement noted in overall or cause-specific survival for men presenting with metastatic prostate cancer over a 20-year period. Cancer 2014;120:818-23. Crossref

22. So WK, Choi KC, Tang WP, et al. Uptake of prostate cancer screening and associated factors among Chinese men aged 50 or more: a population-based survey. Cancer Biol Med 2014;11:56-63.

23. Baade PD, Youlden DR, Cramb SM, Dunn J, Gardiner RA. Epidemiology of prostate cancer in the Asia-Pacific region. Prostate Int 2013;1:47-58. Crossref

24. Carter HB, Albertsen PC, Barry MJ, et al. Early detection of prostate cancer: AUA Guideline. J Urol 2013;190:419-26. Crossref

25. Basch E, Oliver TK, Vickers A, et al. Screening for prostate cancer with prostate-specific antigen testing: American Society of Clinical Oncology Provisional Clinical Opinion. J Clin Oncol 2012;30:3020-5. Crossref

26. European Association of Urology. Guidelines on prostate cancer. Available from: http://uroweb.org/wp-content/uploads/09-Prostate-Cancer_LRV2-2015.pdf. Accessed 1 Jul 2015.

27. Lilja H, Cronin AM, Dahlin A, et al. Prediction of significant prostate cancer diagnosed 20 to 30 years later with a single measure of prostate-specific antigen at or before age 50. Cancer 2011;117:1210-9. Crossref

28. Vickers AJ, Ulmert D, Sjoberg DD, et al. Strategy for detection of prostate cancer based on relation between prostate specific antigen at age 40-55 and long term risk of metastasis: case-control study. BMJ 2013;346:f2023. Crossref

29. Braun K, Sjoberg DD, Vickers AJ, Lilja H, Bjartell AS. A four-kallikrein panel predicts high-grade cancer on biopsy: independent validation in a community cohort. Eur Urol 2016;69:505-11. Crossref

30. Stattin P, Vickers AJ, Sjoberg DD, et al. Improving the specificity of screening for lethal prostate cancer using prostate-specific antigen and a panel of kallikrein markers: a nested case-control study. Eur Urol 2015;68:207-13. Crossref

31. Ng CF, Ng MT, Yip SK. Urology update 2—Management of uro-oncology and associated urological symptoms. Hong Kong Pract 2009;31:128-32.

32. Ng CF, Chiu PK, Lam NY, Lam HC, Lee KW, Hou SS. The Prostate Health Index in predicting initial prostate biopsy outcomes in Asian men with prostate-specific antigen levels of 4-10 ng/mL. Int Urol Nephrol 2014;46:711-7. Crossref

33. Teoh JY, Yuen SK, Tsu JH, et al. Prostate cancer detection upon transrectal ultrasound-guided biopsy in relation to digital rectal examination and prostate-specific antigen level: what to expect in the Chinese population? Asian J Androl 2015;17:821-5.

Sexual violence refers to sexual activity where consent is not obtained or freely given.1 The World Health Organization defines sexual violence as “any sexual act, attempt to obtain a sexual act, or other act directed against a person’s sexuality using coercion, by any person regardless of their relationship to the victim, in any setting”.2 This includes rape, indecent assault, sexual harassment and threats. According to the statistics of ‘Child abuse, spouse/cohabitant battering and sexual violence cases’ from the Social Welfare Department, the majority (88.4%) of newly reported sexual violence cases in Hong Kong are indecent assault and 8.8% are rape or unlawful sexual intercourse. More than 96% of the victims are females. Approximately 70% of the perpetrators are strangers, and the rest are usually someone known to the victim, such as a family member, friend, lover or ex-lover, co-worker, caregiver, neighbour, or teacher.3 Sexual violence is usually underreported because of fear: fear of physical examination, disclosure of sexual history, repeating the traumatic experience in full detail over and over again, complicated legal procedures, not being believed by others, and being harmed by the perpetrator(s).2

Sexual violence may lead to health consequences such as unwanted pregnancy, sexually transmitted disease (STD) infections, physical trauma, depression, and post-traumatic stress disorder.4 Not all victims will seek medical care, however, because the experience of sexual violence is seen as stigmatising and shameful, with possible extreme social consequences.5 Stigmatisation not only from society but also from health care providers, family, and even the intimate partner is common. This leads to minimal support for the victims who may distance themselves by withdrawing from social activities.6

In November 2000, the Association Concerning Sexual Violence Against Women set up the first one-stop crisis centre in Hong Kong, Rainlily, for the protection of female victims of sexual violence. All the social workers at Rainlily are female and trained to provide counselling and care for victims of sexual assault. They will accompany the victim for medical care, police interviews, legal proceedings, and most importantly, the possibly long and difficult recovery period from the incident.

For many years, Rainlily has worked with the accident and emergency department of Kwong Wah Hospital to provide one-stop service to victims of sexual violence including pregnancy prevention, screening and prevention of STDs, forensic medical examination, psychological support, and reporting to the police if desired by the victim. This avoids recalling and repeating the unpleasant experience for different professionals and hence minimises the need for the victim to psychologically re-live the trauma. Since May 2010, Rainlily has also collaborated with the United Christian Hospital and set up an additional rape crisis centre.

We conducted a retrospective analysis of female victims of sexual assault who were seen at either hospital to evaluate the characteristics of the cases and the victims, the use of alcohol and drugs, involvement of violence, and the outcome of the victims. The data are invaluable for health care professionals and policymakers for improving service provision for victims.

All female sexual assault victims who attended the Kwong Wah Hospital or United Christian Hospital from May 2010 to April 2013 were included in this retrospective study. The sexual assault cases were identified from the special case list of the accident and emergency department of each hospital and the designated gynaecology clinic booking list of the United Christian Hospital. The clinical records were reviewed and the demographics of the victims, time lapse from assault to presentation at hospital, characteristics of the assault, investigations and results, treatment and outcome of the victims were analysed.

At Kwong Wah Hospital, the sexual assault cases were managed and followed up in the accident and emergency department, with referral to gynaecologists if clinically indicated, for example, for unwanted pregnancy. At the United Christian Hospital, cases were initially managed in the accident and emergency department with subsequent follow-up in the gynaecology clinic. At initial presentation, victims were screened for the presence of any infection, including STDs. Emergency contraception was prescribed if necessary. Subsequent follow-up was after 2 weeks, 6 weeks, 3 months, and 6 months to exclude pregnancy, to review investigation results and treat any infection.

The study protocol complied with the good clinical practice of ICH (The International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use). Ethical approval was obtained from Clinical Research Ethics Committee of Hospital Authority.

All statistical analysis of data was performed by PASW Statistics 18, Release Version 18.0.0 (SPSS, Inc, 2009, Chicago [IL], US). For continuous data with a highly skewed distribution such as time from the incident of assault to presentation at the hospital, Mann-Whitney U test was used. The critical level of statistical significance was set at 0.05.

From May 2010 to April 2013, a total of 154 sexual assault victims had attended either hospital; 102 at Kwong Wah Hospital and 52 at United Christian Hospital. The age of victims ranged from 13 to 64 years (mean 24.5 years, median 22 years; Table 1). Most (150 cases; 97.4%) victims were Chinese and four were domestic helpers from other Asian countries. Five (3.2%) victims were mentally disabled and 19 (12.3%) had a history of psychiatric disorder.

The time from the incident of assault to presentation at the hospital ranged from 1 hour to more than 5 months (mean 16 days, median 3 days). Half of the victims (n=77) attended hospital within 3 days of the incident. Approximately half (n=84, 54.5%) of the assailants were strangers (Table 2); the others included friend, internet friend, family member, classmate, colleague, employer, boyfriend, ex-boyfriend, and ex-husband.

The median time from the incident to presentation was 48 hours (interquartile range [IQR], 24-240 hours) for victims with symptoms (except pregnancy), compared with 288 hours for those without (IQR, 48-696 hours) [P<0.001]. Those who were pregnant (median time, 756 hours; IQR, 510-1386 hours) presented later than those who were not (median time, 72 hours; IQR, 24-432 hours) [P<0.001].

In 56 (36.4%) cases, alcohol was involved in the incident. There were 18 (11.7%) cases where drugs were involved, including ketamine, amphetamine, methamphetamine, cocaine, and midazolam. In one victim, multiple drugs were involved. Some victims could not identify which drug they had been given. It had either been added to the victim’s drink, or been given as ‘flu medication’ or an ‘anti-drunk pill’.

There were 133 (86.4%) victims with documented vaginal penetration, of whom 25 had also been exposed to oral penetration, five to anal penetration, and four to all three forms of sexual assault. There were three victims in whom penetration was oral only. The remaining 18 victims had no clear documentation. In only three (1.9%) cases did the assailant use a condom. Verbal threats were reported by six (3.9%) victims, and physical violence with or without verbal threat by 45 (29.2%). Reported physical violence included restraint, strangling, beating, grasping, and biting.

Apart from the incident, 75 (48.7%) victims presented with other associated problems, most (n=44, 28.6%) with pelvic or genitourinary symptoms such as lower abdominal pain, vaginal discharge, or urinary symptoms. There were 17 (11.0%) victims who complained of laceration, contusions, bruises, and pain due to physical violence during the incident. Another 12 (7.8%) victims presented with psychiatric or mood problems: two attempted suicide, one had auditory hallucinations, and the others had mood problems or post-traumatic stress disorder with nightmares and flashbacks. One victim presented with per rectal bleeding due to anal penetration and another presented with recurrent oral ulcers in which oral penetration was involved during the incident. Seven victims had found themselves pregnant before attending the hospital.

Blood testing for hepatitis B surface antigen was performed in 146 victims of whom six (4.1%) were positive. All positive results were obtained within 6 weeks of the sexual assault. Hepatitis B surface antibodies were not present in 85 of 134 victims tested. Hepatitis B immunoglobulin was given to 43 victims and a first dose of hepatitis B vaccination to 52. Only 29 victims completed the course of hepatitis B vaccination, however, and the remainder defaulted from follow-up.

Blood test for syphilis by rapid plasma reagin was positive in one victim and was performed around 4 days after the sexual assault. Treponema pallidum haemagglutination assay was also positive. There was no other positive case in the subsequent screening at 6 weeks and 6 months. A similar result was obtained when testing for anti–hepatitis C virus antibody that was positive in one victim and the test was performed within 1 day of the sexual assault. There was no other positive case identified at subsequent follow-up. Blood tests for anti–human immunodeficiency virus antibody were all negative and a total of 71 victims had negative serology 6 months after the alleged assault.

High vaginal and endocervical swabs were taken for culture and revealed one victim with Trichomonas vaginalis. Urethral, rectal, and throat swabs were taken in selected cases and no infection other than with Candida species was detected. Chlamydia trachomatis was tested by polymerase chain reaction test on a urine sample or endocervical swab in 110 victims, and 12 (10.9%) were positive. Among those with chlamydial infection, four presented with genitourinary symptoms such as perineal pain, vaginal discharge, urinary frequency, and dysuria.

Emergency contraception was provided to 63 of the 77 victims who presented within 3 days of the alleged rape. There were 10 victims who had been prescribed emergency contraception, either by other doctors, or self-prescribed from a pharmacy. Other reasons for not prescribing emergency contraception included a victim with only oral penetration, one victim with a previous hysterectomy, two victims taking reliable regular contraception, and one victim who refused the prescription.

There were a total of 11 pregnancies as a result of the incident, among them one victim had received emergency contraception within the day of assault. The gestational age ranged from 6 weeks to 5 months at presentation. Eight cases underwent termination of pregnancy, one underwent medical evacuation for silent miscarriage, one continued the pregnancy to term, and one defaulted from follow-up with unknown outcome.

Attendance at follow-up was 57.8%, 63.6%, 59.1% and 46.8% at 2 weeks, 6 weeks, 3 months, and 6 months, respectively after the incident. Overall, less than half of the victims completed follow-up care.

Every 2 to 3 minutes, one woman is sexually assaulted somewhere in the world.7 The prevalence of sexual violence differs across populations, but studies have consistently shown there to be underreporting in both developed and developing countries.5 It is believed that Hong Kong is no different. The reported cases may only be the tip of the iceberg. This makes prevention, detection, and proper care difficult.

Comparison with a similar study conducted in Hong Kong from 2001 to 2004 by Chu and Tung8 shows that the age of victims, time between assault and presentation, and the percentage of those lost to follow-up are similar; thus the characteristics appear unchanged over the past decade. This raises the question of whether we are doing enough to promote awareness, prevention, and education in society.

A considerable proportion of victims have a history of psychiatric disorder, and there is emerging evidence of the association.9 Although the causal relationship is not well understood, it might be due to the higher prevalence of alcohol or substance abuse among this population. Nonetheless, a history of alcohol or substance abuse was not always documented in the case notes. Patients having a certain type of psychiatric disorder—such as schizophrenia, bipolar disorder, and heroin addiction—are more likely to adopt risky behaviour,10 and hence are at higher risk of being sexually abused.

The delayed presentation among victims of assault by a known assailant may be due to the fear of being discovered by the assailant and being further harmed. Furthermore, sexual violence is associated with stigma in some communities, even in cases with an unknown assailant; the victims may be afraid of being blamed, and there is also a perceived lack of support from families and friends.11 Delayed presentation may result in loss of forensic evidence, delay in prescription of STD prophylaxis, and a missed chance for emergency contraception.12 This explains why those who were pregnant presented later than those who were not, and those without symptoms may not have sought help until they found themselves pregnant. Public education definitely has a role and must emphasise the importance of seeking medical care early and promote community awareness of prevention, instead of blaming the victims.

The prevalence of sexual assault involving alcohol or drug use in this study was similar to a previous study by Hurley et al.13 Females are more vulnerable to the effects of alcohol because of their smaller body mass and higher proportion of body fat. Compared with drugs, there is a higher rate of alcohol involvement in sexual assault cases because it is easily available, cheap, and legally and socially acceptable. Alcohol can cause disinhibition and impair judgement; most of the victims consumed alcohol voluntarily and therefore there is a strong feeling of self-blame after the incident. Recreational drugs consumed by victims themselves or ‘date rape drug’–spiked drinks given to victims can cause sedation, anterograde amnesia, and incapacitation. The actual prevalence is likely to be more than reported, as the victims may not be aware of the assault or only have patchy recall of events. Some of these victims have an intense fear of internet exposure of their body or the incident, and feel a loss of control and sense of insecurity. If a drug abuser is assaulted, they may worry about being charged and are reluctant to report the incident to the police. Delay in reporting or seeking help results in loss of forensic evidence and delay in prescription of emergency contraception. The current protocol of the two studied units did not include toxicological analysis. Therefore the drug used was based on the victim’s report and recall error is highly likely. In order to improve service provision and to help in crime recollection, blood, urine, and nasal swabs for toxicology screening should be obtained in selected sexual assault cases. Public education should emphasise the harmful effects of excessive alcohol consumption and the effects of combining alcohol and recreational drugs. Ways to avoid spiked drinks include keeping an eye on one’s drink, not leaving a drink unattended or obtaining a new one if it is, and not accepting a drink from strangers.

The most common presenting symptoms were pelvic and genitourinary symptoms or injuries as a result of violence during the incident. Psychiatric symptoms were usually underreported. Common symptoms include low mood, fear, guilt, nervousness, sleeping difficulties, poor appetite, and feelings of shame and anger. Emotional numbness and avoidance are common reasons for not seeking help.14 Moreover, some medical providers do not actively ask about psychiatric symptoms. Even if symptoms are reported, they may be considered a ‘normal reaction’ to rape and then ignored. About half of victims recover from acute psychological effects by 12 weeks, but in others the symptoms persist for years.14 Sexual assault survivors are at increased lifetime risk of post-traumatic stress disorder, major depression, suicidal ideas and attempts.15 Mental state and risk of self-harm should be assessed to identify those who are at risk. Psychosocial support and opportunities to talk about the incident are important. For those who do not recover with time, referral to a psychotherapist or even a psychiatrist is essential.

Some experts discourage testing for STD infections in the acute setting unless clinically indicated by symptoms.15 The positive rate of STD in this study was low with the exception of chlamydial infection. Nucleic acid amplification testing can be carried out on urine samples instead of endocervical samples, minimising the need for invasive vaginal examination using a speculum.14 One may consider omitting the screening test and instead offering prophylactic antibiotics against bacterial STDs.

The rate of pregnancy (7.1%) is slightly higher than the quoted risk of 5%.16 The administration of emergency contraception in the two studied units comes in the form of levonorgestrel and was limited to victims who presented within 3 days of alleged rape. Levonorgestrel is licensed for up to 72 hours after unprotected intercourse, indeed there is still some residual efficacy after 3 days although it diminishes with time. To further decrease the chance of pregnancy, other contraceptive methods can be considered in victims who present more than 3 days after the incident, including a copper intrauterine device or ulipristal acetate. If it is not feasible to insert an intrauterine device in the emergency department, urgent referral to a gynaecology clinic should be considered. Ulipristal acetate may not be readily available in all public hospitals but it could be stocked and prescribed as a patient-financed item.

Limitations of this study include the small sample size, recall bias of alcohol or drug use, loss of some victims to follow-up, and short follow-up periods. A territory-wide case review may offer a better evaluation of the problem in Hong Kong.

Sexual assault is usually underreported and can lead to significant health consequences. Involvement of alcohol and drugs is not uncommon in sexual assault cases. Efforts should be made to enhance coordination and cooperation between medical and social services, and improve the accessibility and availability of clinical care. Health care professionals should be properly trained to understand the physical and mental health consequences, the importance of follow-up care, and to equip the skills to manage sexual assault cases. Public education should target at primary prevention, and publicise the simple ways to access the available services.

The authors gratefully acknowledge Mr Edward Choi for his valuable statistical advice.

All authors have disclosed no conflicts of interest.

1. Understanding sexual violence. National Center for Injury Prevention and Control, Centers for Disease Control and Prevention; 2012.

2. Violence against women: preventing intimate partner and sexual violence against women. Geneva: World Health Organization; 2011.

3. Statistics on child abuse, spouse/cohabitant battering and sexual violence cases. Social Welfare Department, the Government of the Hong Kong Special Administrative Region; 2014.

4. Welch J, Mason F. Rape and sexual assault. BMJ 2007;334:1154-8. Crossref

5. Dartnall E, Jewkes R. Sexual violence against women: the scope of the problem. Best Pract Res Clin Obstet Gynaecol 2013;27:3-13. Crossref

6. Jina R, Thomas LS. Health consequences of sexual violence against women. Best Pract Res Clin Obstet Gynaecol 2013;27:15-26. Crossref

7. Masho SW, Odor RK, Adera T. Sexual assault in Virginia: A population-based study. Womens Health Issues 2005;15:157-66. Crossref

8. Chu LC, Tung WK. The clinical outcome of 137 rape victims in Hong Kong. Hong Kong Med J 2005;11:391-6.

9. Goodman LA, Rosenberg SD, Mueser KT, Drake RE. Physical and sexual assault history in women with serious mental illness: prevalence, correlates, treatment, and future research directions. Schizophr Bull 1997;23:685-96. Crossref

10. Hariri AG, Karadag F, Gokalp P, Essizoglu A. Risky sexual behavior among patients in Turkey with bipolar disorder, schizophrenia, and heroin addiction. J Sex Med 2011;9:2284-91. Crossref

11. Abrahams N, Devries K, Watts C, et al. Worldwide prevalence of non-partner sexual violence: a systematic review. Lancet 2014;383:1648-54. Crossref

12. McCall-Hosenfeld JS, Freund KM, Liebschutz JM. Factors associated with sexual assault and time to presentation. Prev Med 2009;48:593-5. Crossref

13. Hurley M, Parker H, Wells DL. The epidemiology of drug facilitated sexual assault. J Clin Forensic Med 2006;13:181-5. Crossref

14. Cybulska B. Immediate medical care after sexual assault. Best Pract Res Clin Obstet Gynaecol 2013;27:141-9. Crossref

15. Linden JA. Clinical practice. Care of the adult patient after sexual assault. N Engl J Med 2011;365:834-41. Crossref

16. Holmes MM, Resnick HS, Kilpatrick DG, Best CL. Rape-related pregnancy: estimates and descriptive characteristics from a national sample of women. Am J Obstet Gynecol 1996;175:320-4. Crossref