Overcoming challenges associated with partner notification following chlamydia and gonorrhoea diagnosis in primary care: a postal survey of doctors and nurses

KEYWORDS: Sexually transmitted infections; contact tracing; partner notification; primary health care; patient care management; chlamydia

Sexual health management guidelines recommend that partner notification (also known as ‘contact tracing’) be carried out when patients are diagnosed with bacterial sexually transmitted infections (STIs) such as Chlamydia trachomatis and Neisseria gonorrhoea.^1,2 Partner notification is an effective case-finding process that has been shown to reduce onward transmission, reinfection and the health consequences associated with new and repeat infections.^3–5 Untreated partners are a known source of reinfection, with an estimated 20% of females becoming reinfected with chlamydia within 6-months of treatment.⁶

While test of cure (within 4–6 weeks of treatment) is not routinely recommended, international and local STI management guidelines recommend routine testing for reinfection at 3 months post-treatment.^1,2,7 Timely detection and treatment of reinfection reduces the more serious reproductive health consequences that can result from repeated infections.⁶ Follow up by phone or text 1 week post-treatment is also recommended to check on treatment compliance and provides an opportune time to check on partner notification progress and offer further advice if needed.^1,2

Limited New Zealand research suggests that most primary care clinicians discuss partner notification, relying on patients to inform their partner(s) about possible infection with no further follow up on outcomes.⁸ A range of barriers to effective partner notification have been identified in research elsewhere.^9–12 This survey of primary care doctors and nurses was conducted to gain an understanding of factors influencing partner notification processes, and to identify strategies that might facilitate more effective partner notification. Views were also sought on new approaches to partner notification that could be trialled in future research.

This study was reviewed and approved by the Southern Health and Disability Ethics Committee on 16 July 2015 (Ref 15/STH/109).

Participants were identified from a laboratory data extract pertaining to chlamydia and gonorrhoea test requests in the Wellington region for the 3-year period 1 July 2012 to 31 June 2015. The dataset included date of test, specimen type, diagnosis, name and location of the requester. Clinicians were eligible for inclusion if they had diagnosed at least four cases of chlamydia or gonorrhoea in 12-months or at least 12 cases in the past 3 years in a primary care setting (general practice, family planning, youth and student health settings). Test requesters from locations not considered primary care (eg tertiary and specialist services including the Sexual Health Service) were excluded. Of the 578 clinicians who had diagnosed at least one case of chlamydia or gonorrhoea in the 12 months to June 2015, 249 met the inclusion criteria. Postal addresses were sought via clinic websites and by phoning clinics, at which point we also checked whether staff still worked at that location. Thirty-three eligible staff were removed at this point (left the practice, retired, maternity leave), reducing our sample to 216 eligible clinicians (147 doctors and 69 nurses).

Survey items were developed following discussion among the research team and review of previously published research,^10–12 and included questions about partner notification practices and experience, and views on suggested barriers and enablers to effective partner notification processes. Survey items were formatted as multi-response, closed-ended questions, with most using Likert scale responses to assess behavioural frequency or levels of agreement with given statements, and free-text comments were invited. The questionnaire was reviewed by four members of the wider research team, piloted with five primary care colleagues, then modified to improve clarity, question relevance and length.

The survey and a cover letter were posted to 216 clinicians in March 2016. We aimed to maximise response rates by checking postal addresses before the mail-out; placing a notice about the research in the main local Primary Health Organisation’s newsletter; offering Continuous Professional Development points for completion, including a personalised cover letter identifying the recipient as part of a group who frequently diagnosed STIs; limiting the survey to one page (two sides) with tick boxes for all questions; including a self-addressed prepaid envelope and a fax number for survey returns, and sending a reminder to non-responders at 4 weeks.

Survey data were entered into Qualtrics software (Qualtrics, Provo, UT, USA) and exported into Excel for collation and analysis. Survey response data were matched using study identifiers (allocated to eligible participants) to our original database of invited participants containing information on clinic type, test and diagnosis volumes. Response frequencies were tabulated for all survey items. Items inviting level of agreement were analysed by combining the responses for ‘agree’ with ‘strongly agree’, and ‘disagree’ with ‘strongly disagree’. Similarly, items asking for estimated frequency of a given behaviour were analysed by combining the responses ‘almost always’ with ‘often/usually’, and ‘rarely’ with ‘never’. For selected items, Chi-square tests compared responses from doctors and nurses.

Completed surveys were returned by 78/147 doctors (mostly general practitioners (GPs)) and 43/69 nurses (mostly practice nurses), giving a total of 121 participants and a 56% response rate. Table 1 depicts the characteristics of the 216 eligible study participants and the 121 who completed a survey.

Table 1. Characteristics of invited study participants and responders

Most respondents felt confident initiating discussion about partner notification (97% agreement, 117/121). Overall, 38% had received training in partner notification (46/121), which was more common among nurses than doctors (49% vs. 32%, P = 0.03). Most thought partner notification should be discussed when results are available or when treatment is given (88.4%, 107/121); 33% (40/121) thought partner notification discussion should occur both at the time of testing and treatment. Of participating doctors, 27% (21/78) reported partner notification discussion was (only) the role of doctors in their practice (4% indicated it was a nurse-only activity). No nurse respondents reported partner notification discussion as solely a nurse’s role, and 16% (7/43) reported it was a doctor-only activity. Seventy-nine percent (34/43) of nurses and 65% (51/78) of doctors indicated it was a shared role (P > 0.05).

In the previous 12 months, 77% (93/121) reported initiating discussion about partner notification with all relevant patients, and 16.5% (20/121) with only some patients. One-quarter of participants (24%, 29/121) agreed that patients in their practice are routinely followed up to check partner(s) have been notified, tested and treated. Approximately, half (52%, 63/121) reported they had followed up with some patients to check on partner notification outcomes, but only 6.6% (8/121) had followed up with all patients.

Figure 1 presents participants’ perceptions of how patients usually respond to discussion about partner notification, with most indicating favourable patient responses. Figure 2 presents responses to statements describing factors that might influence partner notification management. Free-text comments were made by 25 participants at the end of the survey. Selected quotes are presented here with study identifier, clinic type and diagnosis frequency in the past 12 months (ie low, medium or high) included in brackets.

Figure 1. Perceptions of how patients usually respond to discussion about partner notification

Figure 2. Frequency with which suggested factors were perceived to influence partner notification management (provider- and patient-related factors grouped)

One nurse commented:

‘Significant barriers to contact tracing are: transient sexual contacts, sexual partner not having GP service, or simply does not follow up with SHS.’ [NU009, General Practice, Low].

A doctor noted:

‘There are 3 problems. 1. Patients seldom return for follow up due to cost; 2. general practice is becoming overwhelmed by increasing ‘packets’ of unfunded workload and 3. Patients attitudes (and NZ Privacy Act) mitigate against easy discussion of STIs compared with other countries I have worked in.’ [DR047, General Practice, Medium].

Regarding follow up to check on partner notification outcomes, one nurse commented:

‘Young people in our practice are often either a) deliberately vague about their partners or b) don't know how to contact their partners. It is also difficult to follow them up as they have no credit on their phones, change phones, ignore phones and have requested no contact at home.’ [NU019, Youth Clinic, High]

Two participants said that there had to be an element of personal responsibility on the patients’ part to take control of their own health. For example, one doctor commented:

‘We don’t follow up on partner notification - we offer to contact partners, but if the patient says they’ll do it themselves we trust them to do so.’ [DR040, General Practice, Medium]

Figure 3 presents participants’ level of agreement with suggested measures that might improve partner notification and follow up in their practice. We also described a scenario in which a specialist sexual Health Advisor could work with the practice team, taking responsibility for contacting patients to further discuss and counsel patients about partner notification, with phone follow up to check on notification outcomes. Most (82%, 99/121) agreed this assistance would be helpful for some and 45% (55/121) agreed that it would be helpful for all of their patients. A small number – 12% (15/121) – reported that partner notification was currently well managed in their practice, so assistance from a Health Advisor would not be helpful. When asked which of two new approaches they would prefer if involved in future research, one-third (32%, 39/121) indicated they would prefer upskilling their own team to better manage partner notification, 13% (16/121) would prefer working with a Health Advisor and half (49%, 59/121) would be willing to try either approach.

Figure 3. Level of agreement by participating health professionals with suggested measures to facilitate partner notification management

Active recall of patients to test for reinfection (eg via letter or SMS reminder) was uncommon, with only 17.4% (21/121) of participants reporting this almost always or usually occurred, and a further 24% (29/121) reporting that it sometimes occurred. Opportunistic retesting within the next 12-months was reported by 47% (57/121) as almost always/usually occurring, and a further 31.4% (38/121) as sometimes occurring. Participants working in general practice settings were less likely to report this as usual practice (34%, 27/79) than participants working in other settings (71%, 30/42, P < 0.001). Comments made by several respondents highlighted apparent confusion between the recommended 3-month test of reinfection and the less commonly required ‘test of cure.’ For example, alongside questions regarding testing for reinfection, doctors commented:

‘Not recommended in GP land!’ [DR067, General Practice, High];

‘Understood not required unless pregnant.’ [DR010, General Practice, High]

Similarly, a nurse commented:

‘Contradictory advice re. necessity for test of cure, labs not enthusiastic - significant cost also to health service ($85 for chlamydia).’ [NU009, General Practice, Low]

This survey identified challenges in following the recommended best practice for the management of partner notification in primary care. Respondents reported variable levels of discussion with index patients (including absence of discussion), as well as low levels of follow up on partner notification outcomes. Provider- and patient-related factors that most frequently impacted on the management of partner notification in primary care were identified, together with strategies that could facilitate more effective management. Findings also highlighted apparent confusion for some between guideline recommendations to test for reinfection as opposed to ‘tests of cure’.

Participants thought that patients were generally open to partner notification discussion, but appreciated that this sometimes raised difficult issues for their patients. Time constraints were frequently cited as a factor limiting the initial discussion of partner notification, as well as the ability to follow up on outcomes (commensurate with past research).^9,10,12 Investing time in effective partner notification strategies at a community level has real potential to permanently reduce the prevalence of chlamydia and gonorrhoea, and thus provider workload, as well as subsequent patient distress and health consequences.^13,14 Consistent with a survey of Australian GPs,¹¹ factors deemed likely to facilitate partner notification discussion included readily available credible patient resources, training to upskill team members, guidance on what to say and record, as well as access to external advice and assistance when needed. Recognising that discussion with partners about an STI may be challenging for some patients, provision of clear guidance on what partners need to know, together with resources for patients and their partner(s), is likely to assist with this process.

Nearly half of the respondents agreed that if chlamydia and gonorrhoea were notifiable and contact tracing was mandatory, it would be easier to carry out partner notification; and nearly two-thirds agreed it would be helpful having a dedicated external person to assist and have access to expert advice when challenges arise. Gonorrhoea is an infectious disease notifiable to a Medical Officer of Health in the Health (Protection) Amendment Act 2016 pertaining to STIs (effective early 2017), but chlamydia is a non-notifiable disease. While the Act strengthens the legislative provisions for contact tracing, which may be a positive step towards gonorrhoea control, details about how contact tracing would be approached and by whom are yet to emerge.

The volume of opportunistic testing for bacterial STIs and rate of positive diagnoses was considerable for the period reviewed in the process of identifying potential survey participants, reflecting the high rates of STIs in New Zealand.¹⁵ Reinfection rates are known to be high among index patients, and can have more serious reproductive health consequences if undetected,⁶ yet testing for reinfection was not routine practice for most surveyed. Together with comments made by some respondents, this finding suggested a general lack of awareness of retesting guidelines relating to ‘test of reinfection’ (see Table 2 for clarification).

Table 2. Recommendations for clinical practice

We surveyed a relatively small number of primary care clinicians from the large number who diagnose STIs annually, but our sample represented just over half of the clinicians diagnosing the greatest proportion of infections in the Wellington region. We adopted strategies known to maximise response rates,¹⁶ and achieved a 56% participation rate, reaching clinicians from 76% of settings surveyed. Our response rate was lower than hoped for, but exceeded rates reported in similar survey research.^11,17 We selected participants from laboratory test requests, so nurses requesting tests under the name of GP colleagues will have been missed (nurses trained and working within the scope of sexual health practice in New Zealand are able to, and encouraged to register as providers). Views shared by this group of participants might not be representative of the wider group of diagnosing clinicians, with the possibility that our respondents were more likely to more actively engage in partner notification or face the most challenges with this activity. We did not ask about the nature or extent of advice given about partner notification, but presume there would be a range of approaches from brief advice to more intensive motivational interviewing with education and follow up.

Partner notification is an essential part of patient care and STI control. This survey confirms that while partner notification is typically on the agenda following an STI diagnosis in primary care, health professionals (and their patients) need further support to more effectively manage this process. While some of the recognised challenges are outside the control of primary health-care professionals, several strategies could be readily adopted to ensure partner notification is more efficiently discussed (and recorded) within the time constraints of consultations. Table 2 presents a summary of our key recommendations for clinical practice based on these survey findings. These practical but realistic changes to the way partner notification is managed within practice teams could help to ensure that this first step in the process is delivered in a way that maximises the likelihood that patients feel motivated and able to notify their sexual partner(s), and reduce their risk of reinfection. Further research to identify patient perspectives and factors influencing the likelihood that they talk to their partner(s) could help to further improve the effectiveness of this process.

None.