Abstract
Health disparities often reflect inequitable access to appropriate health care. This study aimed to establish if cases of genital chlamydia infection were managed equitably by age, gender and ethnicity in a region of New Zealand with high rates of chlamydia infection (858 per 100,000 population). Clinical records of 415 genital chlamydia cases from 19 different health-care sites, including general practice and community settings, were reviewed. Data were analysed by demographic variables. For those treated, men were treated more quickly than women (median 3 days versus 6 days, P < 0.001), but there was no difference by ethnicity. Cases without documented treatment were more likely to be women (8.2% versus 2.1%, P = 0.037) and more likely to be Māori than non-Māori (13.6% versus 4.8%, P = 0.036). Overall, the most notable issue was the lack of effective partner notification across all demographic variables. Ongoing efforts are required to ensure equitable access to timely treatment and to ensure that more effective partner notification strategies are implemented.
INTRODUCTION
Chlamydia trachomatis infection (chlamydia) is a significant public health problem, as untreated infection may lead to salpingitis, tubal scarring, ectopic pregnancy and subfertility in some women. 1 New Zealand's estimated national rate of 803 per 100,000 population in 2009 was two to four times higher than reported national rates for Australia (287 per 100,000 population in 2009), the UK (202 per 100,000 population in 2008) and the USA (401 per 100,000 population in 2008). 2 Waikato District Health Board (DHB) in the upper North island of New Zealand has a high burden of reported cases (858 per 100,000 population in 2009), with nearly 80% being among those less than 25 years old. In addition, available data suggest disparities for Māori, with reported sentinel surveillance clinic rates of chlamydial infections among Māori being 2.5 times than that of non-Māori. 2
Disparities in health outcomes often reflect socioeconomic and other factors that impact on adequate access to appropriate health care. 3,4 Recent efforts locally have therefore focused on improving district-wide primary sexual health-care provision for young people, with free general practice visits introduced in high-need areas during 2003–2004. 5 Encouragingly, chlamydia test uptake within Waikato DHB is now equitable by ethnicity, with high rates of testing among both Māori and non-Māori women under 25 years old, 6 and coverage for young women reached 45% in 2009. However, improved access to testing without ensuring equitable access to effective treatment or partner notification means that disparities in health outcomes will likely persist. Hence, our aim was to ascertain if there were any disparities in current case management of genital chlamydia infection in a range of clinical settings in Waikato DHB.
METHODS
Waikato DHB had an estimated resident population of 357,000 in 2008, of whom approximately 21% are Māori compared with 15% nationally. Nearly 54% of Māori are under 25 years of age, compared with 31% of non-Māori. Since mid-January 2008, two Waikato laboratories perform all chlamydia testing for the district. Both laboratories provided data on all tests carried out on residents from 1 February to 31 October 2008. Non-genital site samples were excluded. All urogenital samples were tested using nucleic acid amplification methods. Positive chlamydia test results were identified. Any practice or clinic setting within Waikato DHB with 25 or more positive chlamydia test results during the nine months was invited to participate. Each site was provided with a list of their laboratory-identified cases and asked to complete a proforma for each of 20 consecutive cases seen from 1 February 2008. The New Zealand Ministry of Health guidelines for chlamydia management 7 are closely aligned to the UK guidelines for the management of uncomplicated genital chlamydia infection; hence, a UK-validated national chlamydia management audit tool was used as a basis for our proforma. 8 The proforma was adapted to more closely reflect the New Zealand context and differences in health-care settings (Appendix A).
Twenty sites across a range of clinical settings were eligible. This included: nine rural general practices, three urban general practices, a family planning clinic, a sexual health clinic, a community accident and medical centre, a remand prison, a university-based student health service, high school-based student health services, a hospital-based emergency department and a hospital-based acute gynaecological service. All sites agreed to participate and 19 of 20 were able to provide data. The non-participating site was the remand prison. Combined, these sites detected 70% of 2258 urogenital chlamydia cases diagnosed in Waikato DHB during the time period. Seven sites chose to complete proformas for more than 20 cases (range 21–37), giving a sample of 415 cases (18%) of all Waikato DHB genital chlamydia cases diagnosed during 1 February–31 October 2008.
The analysis with regard to whether recommended standards of care were met has been described previously. 9 This analysis focused on differences in case management by age, gender and ethnicity. Chi-squared and Fisher's exact tests were used to examine demographic and other categorical variables. Kolmogorov–Smirnov testing was used for continuous, but non-normal variables, such as days to treatment. Ethical approval was given for the study (NTY/09/25/EXP).
RESULTS
Of the 415 cases, 316 (76%) were women and 317 (78%) were under 25 years old (Table 1). Two transgender patients were excluded from any gender analysis. Ethnicity data were recorded for fewer laboratory-identified cases than for study cases (82% versus 95%, P < 0.001). The most commonly reported ethnicity among study cases was European (48%), followed by Māori (43%). A range of ethnicities were noted but, because of small numbers for several of these, ethnicity was analysed as Māori and non-Māori. There were more cases from rural general practice and less from sexual and reproductive health providers, P < 0.001.
Demographics of all Waikato DHB cases and of study sample
*Including two transgender cases
Presenting features
The main reason for having the chlamydia test was documented for 319 of 415 cases (Table 2), with ‘symptoms’ being the most frequently reported reason for both sexes. Women were more likely than men to have ‘offered by provider based on sexual history’ as the main reason for testing (P = 0.007) while men were more likely to have ‘contact of partner diagnosed with chlamydia’ as the main reason for testing (P = 0.007). By ethnicity, Māori were more likely than non-Māori to have tests that were ‘offered by provider based on sexual history’ (41% or 24% versus 20% or 9%, P < 0.001).
Case management by gender
*Two transgender cases excluded
Overall, 57% of cases were symptomatic. There was no difference in symptomatic cases versus asymptomatic cases by age-split (57% of under 25s versus 55% of those 25 and over, P = 0.8), by gender (65% of men versus 54% of women, P = 0.06) or by ethnicity (55.6% of Māori versus 55.9% of non-Māori, P = 0.96).
Treatment documentation
Treatment drug, dose and duration were well documented for 90% (373 of 415) of cases. Treatment was noted as being prescribed for an additional seven cases, but there was no record of drug choice or dose. Thus, 92% of cases had some evidence of treatment. Where treatment was documented, drug choice, dose and duration of treatment were appropriate for 98% of cases when compared with the New Zealand Ministry of Health guidelines for chlamydia management. 7 Treatment, where noted, was mostly by prescription, rather than direct dispensing (73% versus 27%), with no difference in this by demographic variables.
Time to treatment was recorded for 90% (373 of 415) of cases. Same-day treatment occurred for 23% (85 of 373), 90% were treated within 11 days and 95% within 17 days of being tested. The remaining 5% (18 of 373) were treated between days 19 and 90 of being tested. The only demographic difference in time to treatment was by gender. There was significantly less time from test to treatment for men than for women (median 3 days versus 6 days, P < 0.001). Only 16% (45/278) of women with documented treatment were treated at the time of testing, compared with 42% (40/95) of men, P < 0.001.
When analysed by reason for testing, known contacts had the shortest median time-to-treatment (1.3 days), while provider-offered testing had the longest median time-to-treatment (7.5 days). Those whose main reason for testing was ‘offered by provider’ was also less likely than those requesting a check-up to have documented treatments (78% versus 99%, P < 0.001).
No treatment was recorded for 35 cases; five cases failed to attend for treatment and the remaining 30 cases had no documented treatment or outcome, although it is possible that treatment occurred. Of cases with known ethnicity, lack of documented treatment or failure to attend for treatment was more likely for Māori than non-Māori (13.6% versus 4.8%, P = 0.036). Lack of documented treatment was also more likely for women, with 8.2% not having any treatment documented, compared with 2.1% for men, P = 0.037. There was no difference by age.
Lack of documentation for treatment or outcomes was notable in cases diagnosed in two hospital-based settings, with 26 of 47 (55%) of cases having incomplete or no treatment documented. In comparison, 16 of 368 (4.3%) cases diagnosed in other settings had either incomplete or no treatment documented. Of cases with known ethnicity, more Māori were seen at these two settings than elsewhere (74% versus 38%, P < 0.001). Also, more women than men attended these two settings (43 of 47 cases, 91%, P = 0.01).
Partner management
The results of partner notification compared with the Ministry of Health guidelines for chlamydia management 7 have been described previously. 9 To summarize, most cases had limited documentation in their medical record around partner notification; 59% (246 of 415) had any documentation to suggest partner notification was discussed and 31% had any documentation about the planned method of partner notification (e.g. by the patient). Most participating sites noted that discussions typically involved telling patients to tell their partners. Documentation that partner notification was discussed did not differ significantly by age-split, gender or ethnicity.
A recommended target for partner notification is at least 50% of identified sexual partners being treated. 7 Detailed sexual history taking appeared to be lacking; only 47% (196 of 415) of cases had any indication of the number of sexual partners in the preceding three months. There was also limited documentation by which to monitor the success of partner notification; 21% cases had any documentation that regular sexual partner(s) had been advised, 12% noted regular partner treatment as advised by the index case and 4% noted provider-verified regular partner treatment. There were no significant differences within any of the measures with respect to gender, age-split and ethnicity.
DISCUSSION
The limited partner notification, whether by age, gender or ethnicity, noted in a wide range of clinical settings in our locale is a concern. Repeated chlamydia infections in young women are most often re-infections and frequent testing and treatment of women alone is not likely to reduce prevalence in high-risk populations. 10 Documentation that partner notification was discussed was similar to that reported in other community settings. 11,12 However, with the exception of the Sexual Health Clinic, there was little patient follow-up and documented outcomes of partner notification were notably lower than that reported by the UK settings, 12–14 where partner notification is most often undertaken by specialist sexual health advisers. 15 In New Zealand, most sexually transmitted infections are managed in primary care and very few health boards employ contact tracers or health advisers for sexually transmitted infections. Improvements are feasible. With training, practice nurses can undertake partner notification that is at least as effective as referral to a specialist health adviser. 16 New methods of partner management are emerging and a systematic review of interventions found involving index patients in shared responsibility for the management of sexual partners improved outcomes, for example, patient delivered partner therapy, home sampling for partners and providing additional information for partners. 17
Of diagnosed cases in this study, men were no more likely to be symptomatic than women, but more men were treated immediately. This is presumably because male urethral symptoms, particularly urethral discharge, more accurately predict sexually transmitted infections than female symptoms, such as vaginal discharge. 18 Men in the study were also more likely to be a contact of a chlamydia case. Others have reported men as being more likely to receive immediate treatment, 19,20 and have suggested clinical features or the use of on-site microscopy as important influences. None of the settings involved in our study use on-site microscopy yet, encouragingly, the overall median time-to-treatment of five days is comparable to that previously reported from specialist genitourinary clinic audits. 21,22 However, for women in our study, the median time-to-treatment was double that of men and they were also more likely to have a lack of documented treatment. If treatment is not given immediately, a range of structural health-care factors, including the rapidity with which tests results are received, influence the time to treatment. 20,23 Having a reliable point of care test to enable more immediate treatment for women would help overcome these factors.
Analysis by reason for testing suggests those that request testing may be more likely to have documented treatment than cases where testing was provider-offered. This raises the issue of whether provider-offered testing may be associated with a lower likelihood for returning for results and/or treatment, and warrants further investigation. Also, compared with women, case finding for men appeared more related to clinical features, such as symptoms or being a contact of infection, and less related to provider-offered opportunistic testing. This raises the possibility of missed opportunities to offer chlamydia testing to young men, as was noted in a UK general practice study. 24 New Zealand has achieved high chlamydia test uptake among young women, but not young men, with five times as many women aged 15–24 years being tested. 6,25 However, our data does not capture if men were more likely than women to receive syndromic treatment without testing for chlamydia.
New Zealand's ethnic disparities in access to health care are well documented. 3 There may also be disparities in the quality of health care. 26 In our study, treatment outcomes, such as time-to-treatment, appropriateness of treatment and partner management, were similar. However, there were some differences for Māori compared with non-Māori. Reasons for testing among Māori were more likely because of provider-offered testing than by patient-requested testing. It may be that providers perceive Māori as being at greater risk of having chlamydia and hence be more likely to offer testing as, although local test uptake is equitable by ethnicity, Māori are twice as likely as non-Māori to test positive. 6
Lack of documented treatment or failure to attend for treatment was also more likely for Māori than non-Māori; although the numbers affected were small, this occurred for more than one of every seven Māori compared with less than one in 20 non-Māori. This may be another measure of access, as Māori may be more likely to face structural barriers in returning for treatment, e.g. poor transportation. However, this finding may simply reflect the bias around documentation for some settings involved in this study. As reported in the overall audit, secondary-care test results were often available only after a patient had been discharged from hospital and it was not possible to ascertain from the hospital medical record that cases and their partners were ever treated. 9 Those facing cost barriers to health care may be more likely to choose to attend the local emergency department, particularly out-of-hours when there are limited options for free or low-cost primary care in Waikato DHB. This study was not designed to look at these issues and further study is needed to clarify this.
Study limitations include the retrospective study design and purposive sampling of cases. Only settings that diagnosed 25 or more cases were invited to participate, so that there was less inherent bias related to small case numbers. Relatively more cases were from rurally based general practice and less from sexual and reproductive health services, meaning that these results are not generalizable to settings where sexual health care is delivered predominately at specialist clinics. Data collection was based on the clinical record where documentation may not reflect actual care. Some pertinent information was not available, for example, 73% of treatment was prescribed rather than dispensed, but there were no data regarding adherence to non-dispensed medications. Study strengths include a high participation rate and measuring clinical performance in a wide-range of settings. Ethnicity data for the study were collected from the patient medical record. Historically, ethnicity data within New Zealand databases have been provider-assigned and self-identified non-European ethnicities undercounted as a result. 27 However, local health-care settings have made efforts over recent years to collect self-identified ethnicity within their patient management systems. This information was felt to be more accurate than National Health Identifier (NHI)-mapped ethnicity within the laboratory database, where 18% of cases had missing data.
In conclusion, our study found differences in case management by gender and ethnicity that suggest barriers to timely appropriate health-care persist. However, by far the most notable issue was the overall lack of effective partner notification and, until this improves, prevalence is unlikely to reduce and current disparities in reported infections will likely continue. Ongoing efforts are required to ensure equitable access to timely treatment and to ensure more effective partner notification strategies are implemented.