Partner notification services of the Sexually Transmitted Disease clinics in Sri Lanka

Introduction: Partner notification is a strategy used to control sexually transmitted infections (STI). It includes identifying a look-back interval and offering testing and treatment to the relevant sexual partners of an index patient. Objective: Study was carried out to assess partner notification (PN) services at selected Sexually Transmitted Disease (STD) clinics. Method: Study consisted of four components. Component 1 was a cross sectional descriptive study among the staff of five selected STD clinics who were directly involved in PN, to assess knowledge, attitudes and practices on PN. Component 2 was a retrospective descriptive study in the same setting to describe the burden of partner notifiable STIs during 2012. Component 3 was a casecontrol study among patients with partner notifiable STIs in the STD clinic, Colombo during 2012 to describe factors associated with patient compliance in PN. Component 4 was a cross sectional study at STD clinic, Colombo to evaluate PN indices. Results: All interviewed staff knew that gonorrhoea, early syphilis and chlamydia infection are partner notifiable but some misidentified bacterial vaginosis (20%) and vaginal candidiasis (14%) as partner notifiable. Knowledge on contact actions and look back period were unsatisfactory and 17% and 10% respectively were unaware that look back period is irrelevant to genital herpes and genital warts. A total of 1,157 partner notifiable STIs have been reported from all five clinics in 2012. Patient compliance in PN was significantly associated with regular partners, NGI and coitarche ≥ 19 years. Percentages of index cases interviewed for PN and index cases who had documented outcomes for all contacts were only 30.7% and 8.1% respectively. Conclusions: Knowledge, attitude and practices on PN among the clinic staff was unsatisfactory. Percentage of index cases interviewed was low (30%) probably leaving a considerable number of contacts unattended. Patient compliance on PN was significantly associated with regular partners, coitarche ≥ 19 years and NGI. Documentation relevant to PN was unsatisfactory.


Introduction
Partner notification (contact tracing) is a strategy used to control STIs. It involves identifying a look-back interval in which infection of contacts may have occurred, informing the relevant contacts of their exposure, offering testing and treatment and recording contact actions and outcomes. (1) Following contact actions are used in partner notification. Patient referral: Patients are encouraged to notify partners at risk. Provider referral: Health care workers assist in notifying partners. Contract referral: A time frame is negotiated by the provider, for the patient to notify partners. (2) No action: This is appropriate when a contact is non-traceable or has been verified as already visited a health-care facility. (1) Gonorrhoea, syphilis, NGI and trichomoniasis were considered as partner notifiable infections (3) in this study.

Partner notification process in Sri Lanka:
The medical officer (MO) refers the relevant patient to the public health inspector (PHI) or the public health nursing sister (PHNS) for partner notification interview to identify primary contacts of the index patient.
All possible details of the contacts are recorded in  Patient's record  Interview and contact tracing register  Counselling and partner notification form (H-1205) Contact action is determined with the informed consent of the index patient. Contact slips are issued (3) if patient referral is the agreed action.
This study was performed to understand partner notification among patients with STIs in Sri Lanka since no such study has been done so far STD clinics at Colombo, Ragama, Gampaha, Kalutara and Kandy were selected as study setting since it was assumed that Patient Information Management System (PIMS) implemented in those, will enable gathering data more systematically.
The objective of the study was to assess the level of partner notification in selected STD clinics and to understand the knowledge, attitudes and practices of health care workers relevant to partner notification and factors associated with compliance with partner notification.

Methods
The study had four components with different study designs as outlined below. (Table 1) Data collection from the clinic staff (Component 1) was done using a structured self-administered questionnaire which included questions on their sociodemographic information, basic and essential knowledge, attitudes and practices on partner notification.
Data on partner notifiable STIs (Component 2) were extracted from the PIMS.
Data for the case control part (Component 3) was obtained by a data extraction sheet structured to extract data from PIMS, patients' records, H-1205 and interview and contact tracing register.
Registration numbers of new patients with partner notifiable STIs were obtained from main register and the data were collected using the data extraction sheet.
Data collection on partner notification of the index patients attended during 2012 was continued for further six months (till the 31 st of June 2013). It was assumed that all reasonable steps to notify partners have been taken by that time.
Evaluation of partner notification services at STD clinic, Colombo (component 3) was done using the same data extraction sheet developed for the component 2. In addition, data on treatment status of the contact was gathered from the clinic record of the contact. Date of data entry in to H-1205 was considered as the date of interview of the index patient as there was no separate place to enter the date of interview. This was important in calculating the time taken for the contact to come to the relevant STD clinic following partner notification interview.
Patients were considered as interviewed, when the particular patient's clinic record contained filled (H-1205). There were 229 interviewed patients. Cases and controls were selected out of those as below.
Cases: Interviewed index patients of whom at least one partner attended the clinic Controls: Interviewed index patients of whom at least one partner did not attend the clinic Following formula was used to calculate the sample size for the case control analysis. (4) P 1 (proportion of exposure among cases) = 26% P 2 (proportion of exposure among controls) = 9 %. r= 2 (ratio of cases: controls = 1:2) Level of significance = 0.05.
The number of cases (n) was calculated as 57 and 5% added for missing data. Final sample size was 60. Since the ratio of cases to controls was taken as 1:2, number of controls needed was 120. However, there were 169 patients with partner notifiable STIs who met the eligibility criteria as controls and they all were included in the analysis to further increase the power of the statistical analysis. Data were entered in to the SPSS version 13 for descriptive statistical analysis.

Results
Participants (22 MOs and 10 public health staff officers) represented 70% of all staff attached to these five clinics. Majority were males. Over 90% were in the age group of 30-53 years, and over half of them had 1-5years of service experience (Table 2) The table 3 contains the questions and the proportions of staff that gave the correct answers.   condom use in last 3 months, condom use at last sexual exposure and sexuality were not significantly associated with the compliance with partner notification. Table 5 indicates the variables that had a significant association. Since the significance could have been due to confounding effects, binary logistic regression analysis was done. Significant variables given in table 5 and four other likely important variables were included in binary logistic model. Table 6 shows the last step of the binary logistic regression analysis done using backward logistic regression option. Following four variables were statistically significantly associated with the compliance with partner notification.

Discussion
Despite the mandatory training given to all concerned staff at the beginning of their carrier in a STD clinic and reasonably long service period, basic knowledge of partner notification among them was not up to an acceptable level.
All interviewed staff knew that gonorrhoea, early syphilis and chlamydia infection are partner notifiable but some misidentified bacterial vaginosis (20%) and vaginal candidiasis (14%) as partner notifiable. Only61.3%were aware that late syphilis is partner notifiable. Knowledge on look back period was unsatisfactory and 17% and 10% respectively were unaware that look back period is irrelevant to genital herpes and genital warts. Only nearly 1/3 knew the correct look back period for late syphilis.
Some participants (6.5%) have stated that partner notification is not important in bacterial infections. More than half of the sample has identified limited time and manpower as the main problems they encounter in partner notification in their clinics.
Prevalence of syphilis among males was nearly twice as that of females while the prevalence of NGI among males was nearly half as that among females. In contrast, the WHO estimate for the region (5) states that the prevalence of chlamydia infections is nearly twice among males (chlamydia is considered to cause 70% of non gonococcal urethritis) as that of females but the estimation for the prevalence for syphilis is the same for both sexes. The higher prevalence of NGI among females in the present study can be explained by the mostly subjective diagnosis among females in contrast to the objective diagnosis among males in Sri Lanka.
Although it is reasonable to anticipate an inverse relationship between the prevalence of STIs and the educational level, it was striking to observe that both groups of patients who had no schooling and who had diploma or degree had almost the same proportion (4%) of STIs.
Being married or co-habitant was significantly associated with compliance with partner notification. The reasons for this were not assessed in the present study. A systematic literature review in Africa, Asia and Latin American and Caribbean countries has revealed that major barriers for notifying casual and commercial partners were nontraceability of them and that men had no further plans to continue sexual relationships with them. (6) In this study, gonorrhoea and syphilis did not have an association with compliance. But the patients who had NGI were 2.5 times less likely to be compliant with partner notification than those who did not have NGI.
Last sexual exposure within four weeks was significantly associated with compliance, after univariate analysis but after binary logistic regression that association was not found to be significant. An observational study among patients with gonorrhoea or Chlamydia infection in Netherlands indicated that last exposure less than 1 week has been significantly associated with the success of partner notification Vs last exposure more than 1 month back. (7) Percentage of index cases that had at least one interview for partner notification was 30.7%.In an audit done by the principal investigator in a sexual health clinic in Portsmouth, UK (8)  British statement mentions that outcomes of all agreed contact actions should be completed by 28 days. Anyhow in this study, six months were allowed after the first interview, for partners to be brought before closing each scenario but allowing a longer period did not make a better outcome.
Although the values obtained in this study cannot be compared with the British standards or the values from any developed country, these results are below satisfactory level.

Conclusions
The level of Knowledge, attitude and practices on partner notification among clinic staff was unsatisfactory. The percentage of index cases interviewed was low (30%) probably leaving a significant number of contacts unattended. Patient compliance on partner notification was significantly associated with being married/co-habitant, last sexual exposure with regular partner, coitarche ≥ 19 years and NGI. Documentation relevant to partner notification was unsatisfactory. It is advisable to well train the staff on partner notification and to introduce one format to record all data in partner notification instead of using many formats. Additional information such as number of partners elicited, agreed contact action, status of notification, detailed outcome of an agreed contact action should be included in the new format. Provider or contract referral was never been observed within the study. More attention should be paid to improve those contact actions by providing the STD clinics with essential resources such as transport facility and funding.