|Home | About | Journals | Submit | Contact Us | Français|
To characterise patients who decline to provide their surname at a public sexual health centre.
A case–control study of all patients first attending the Sydney Sexual Health Centre from 1998 to 2004, using proforma‐collected electronic data to compare patients who did not provide their surname with those who did. In addition, the frequencies of the 10 most common surnames in the Sydney telephone directory were compared with the frequency of those names in the patient database.
Of 27241 patients, 1350 (5%) declined to provide their surname. The most common surnames were also over‐represented, suggesting that aliases remained pervasive among the centre's patients. Sex workers, married people and people requesting HIV, hepatitis or sexually transmissible infection (STI) screening were all more likely to decline to provide a surname. By contrast, patients with symptoms, patients who were referred with a prior STI diagnosis and patients with a new bacterial or non‐HIV viral STI or were a known contact with STI were all significantly more likely to provide a surname. Among patients who declined to provide a surname, 20 tested HIV positive.
The anonymous option did not seem to eliminate the use of aliases. Although limited, there seems to be a market for anonymous sexual health screening, particularly for the asymptomatic.
Confidentiality concerns may act as a barrier to accessing healthcare for people at risk of sexually transmissible infections (STIs), including HIV.1 For this reason, public sexual health services worldwide allow patients to register without proof of identity, often resulting in incomplete or false information being provided. From a clinical perspective, working with patients who are anonymous or are using aliases can be problematic, particularly when care is shared with other agencies.
To justify the policy of allowing patient anonymity, it would be useful to determine if the policy conferred a public health advantage. Studies in the US have produced conflicting results about the importance of anonymity for those seeking HIV testing.2,3 Thus, we sought to characterise patients who declined to provide a surname at an inner‐city public sexual health service.
Since 1991, as an alternative to the previous practice of inviting false names, the Sydney Sexual Health Centre, Sydney, New South Wales, Australia, has explicitly allowed patients to register without providing a surname by simply entering the patients' surname as “declined” into the patients' record and the centre's database if the patient exercised this option. This has the advantage of reducing the number of patients using (often forgotten) false names at each visit. The files of these “declined” patients can be retrieved by searching other demographic variables such as date of birth. For this study, data on all new patients between 1 January 1998 and 31 December 2004 were extracted from the centre's database. The “declined” patients were compared with the rest of the clinic population for demographics, risk behaviour, reasons for attendance, STI testing rates, and current and future diagnoses.
In addition, to determine whether some patients may have provided a false surname despite having the option of providing none, the frequencies of the most common surnames in the Sydney telephone directory4 were reviewed to see if these surnames were over‐represented among the centre's patients. This exercise was based on the assumption that people tend to use a more common name if adopting a false identity. The local Area Health Service provided ethical clearance for the study.
Data were analysed with Stata V.8.0 for Windows. We modelled associations by univariate and multivariate logistic regression analyses, and used Wald tests and related confidence intervals (CIs) to assess statistical significance and precision.
Of 27241 new patients who attended the centre between 1998 and 2004, 1350 (5%) overtly declined to give a surname (the “declined” group). This “declined” rate did not vary significantly during the study period (data not shown). The “declined” group were older (mean age 32.3 v 29.8 years, p<0.001) than the 25891 patients who provided a surname. Table 11 details the other characteristics of the “declined” group. At the univariate level, the “declined” group were more likely to be women, but after controlling for sex work and other variables, multivariate analysis showed no difference in sex. As none of the 28 transgender patients declined to provide a surname, they were excluded from further analysis.
Reporting a current (adjusted odds ratio (AOR) 1.59) or previous (AOR 1.55) marriage or defacto relationship or—more potently (AOR 5.47)—failing to report marital status were predictive of patients being in the “declined” group compared with those who had never been married (table 11).). Exclusively heterosexual men were less likely to be in the “declined” group at univariate analysis but significantly more likely (AOR 1.33; 95% CI 1.14 to 1.54) at multivariate analysis.
The reasons for patients' presentation to the centre and the current diagnoses were not mutually exclusive, so these characteristics were not amenable to multivariate analysis. At univariate analysis, patients reporting symptoms (OR 0.75), known contact with a STI (OR 0.6) or management of a previously diagnosed condition (OR 0.46) were significantly less likely to be in the “declined” group. By contrast, patients requesting screening for HIV, hepatitis or STIs were more likely to be in the “declined” group (table 11).
Patients, male or female, who reported any same sex partners in the past 12 months were more likely to be in the “declined” group (AOR 1.49), although the reported number of non‐paying sexual partners in the previous 3 months or a history of injecting drugs had no association (table 11).). Current engagement in sex work was the strongest single predictor of a patient being in the “declined” group (AOR 5.68). These 329 sex workers comprised 24.4% of the total “declined” group.
Perhaps because they were more likely to be symptomatic, patients diagnosed with a current bacterial or non‐HIV viral STI were significantly less likely to be in the “declined” group.
Being initially in the “declined” group was not associated with being diagnosed during a later episode with HIV (OR 1.28; 95% CI 0.8 to 2.05) or a bacterial STI (OR 1.18; 95% CI 0.99 to 1.42; data not shown).
When the frequencies of the 10 most commonly used surnames in the Sydney telephone directory4 were compared with the frequency of surnames used by patients at the centre (table 22),), the most common names were found to be moderately over‐represented among the centre's patients. One name—the Vietnamese family name of Nguyen—was obviously under‐represented. Sydney's two concentrations of Vietnamese Australians are located 10 and 30 km away and are served by local sexual health clinics.
We found that 1 of the 20 patients at this inner‐city sexual health clinic declined to provide a surname when they were explicitly given the option. Additionally, the most common surnames in the telephone book were over‐represented among the centre's patients when they did provide a surname. These findings indicate that large minorities of public sexual health patients have a desire for anonymity even within this confidential inner‐city clinical setting.
To place these findings in context, most STI and HIV testing and management in Australia occurs in general practice,5 wherein anonymity is not a practical option because costs have to be claimed on a named basis from Australia's universal health insurance scheme, Medicare. Thus, public clinics would tend to attract disproportionate numbers of patients who may have a heightened desire for anonymity, as well as those who are ineligible to claim medical costs against Medicare.
All of the transgendered people gave a surname at registration. As many transgendered people socially function with an adopted name, this may simply reflect their comfort with using an alias.
Current sex work was the strongest predictor of being in the “declined” group. Anecdotally, many sex workers who do provide a surname at the centre use their adopted “working” name rather than their birth name: an observation that we were unable to quantify. Although sex work is decriminalised in Sydney6 it remains socially stigmatised. Confidentiality concerns may be further magnified for migrant sex workers.7
Marriage or defacto status, current, past or undisclosed, also predicted a patient being in the “declined” group. As >95% of other demographic variables are available in the patient database, so many patients failing to disclose their marital status and being in the “declined” group suggests that most were married. We interpret this as indicating a heightened demand for anonymity for social and relationship reasons.
Interestingly, the presence of genital symptoms, known STI contact, referral for management of a previously diagnosed STI, or a new diagnosis of a bacterial or non‐HIV viral STI were all associated with a reduced likelihood of being in the “declined” group. This can be interpreted as an example of clinical imperatives over‐riding the patients' confidentiality concerns. Many would have been referred by other services where anonymity was not an option. In our experience, patients who are familiar with our service readily revert to using their full names if they need to be referred elsewhere or if they require treatment for their HIV infection. Subsidised drugs for HIV mandate the patient producing a Medicare identity card.
By contrast, being a patient who is requesting HIV, hepatitis or STI screening—or accepting HIV testing when offered—was associated with being in the “declined” group, suggesting an ongoing concern about issues of confidentiality in less clinically driven patients. It is not known whether these patients would have refused testing for HIV if they were not given the option of anonymity or whether more would simply have given false names. Over the 7‐year study period, a total of 20 patients in the “declined” group tested positive for HIV either at the initial visit or during a subsequent episode, a yield similar to patients who gave a surname (table 11;; p=0.48).
From a public health perspective, the patients in the “declined” group in our clinic were not conspicuously at higher or lower risk than the patients who provided a surname. However, in settings where sex workers have higher STI rates, offering an anonymous option could have a marked public health benefit. In many settings, sex workers are disproportionately affected by HIV/AIDS, and healthcare services need to adapt to their needs.8 Although limited, there seems to be a market for anonymous sexual health screening, particularly for the asymptomatic.
AOR - adjusted odds ratio
STI - sexually transmissible infection
Competing interests: None.