Gonorrhoea and Syphilis Epidemiology in Flemish General Practice 2009–2013: Results from a Registry-based Retrospective Cohort Study Compared with Mandatory Notification

Background The number of newly diagnosed gonorrhoea and syphilis cases has increased in Flanders in recent years. Our aim was to investigate, to which extent these diagnoses were registered by general practitioners (GPs), and to examine opportunities and limits of the Intego database in this regard. Methods Data from a retrospective cohort study based on the Flemish Intego general practice database was analyzed for the years 2009–2013. Case definitions were applied. Due to small case numbers obtained, cases were pooled and averaged over the observation period. Frequencies were compared with those calculated from figures of mandatory notification. Results A total of 91 gonorrhoea and 23 syphilis cases were registered. The average Intego annual frequency of gonorrhoea cases obtained was 11.9 (95% Poisson confidence interval (CI) 9.6; 14.7) per 100,000 population, and for syphilis 3.0 (CI 1.9; 4.5), respectively, while mandatory notification was calculated at 14.0 (CI: 13.6, 14.4) and 7.0 (CI: 6.7, 7.3), respectively. Conclusion In spite of limitations such as small numbers and different case definitions, comparison with mandatory notification suggests that the GP was involved in the large majority of gonorrhoea cases, while the majority of new syphilis cases did not come to the knowledge of the GP.


Introduction
In the last decade, the number of newly diagnosed sexually transmitted infections (STIs) has risen substantially in Flanders. There, the number of new gonorrhoea cases obtained by mandatory notification increased from 454 in 2006 to a maximum of 1590 in 2015, and the number of syphilis cases from 312 in 2006 to a maximum of 663 cases in 2015, respectively [1].
In 2007, Verhoeven and colleagues deplored the lack of available -reliable data on the importance of the general practice setting in STI diagnoses‖ [2]. A study on East and West Flanders in 2010 [3] and the follow-up study in 2012-2014 [4] showed that GPs treated nearly 80% of gonorrhoea patients in the study. GPs are the largest group of primary care practitioners, but the extent of their involvement in STI control in Flanders is still incompletely understood. Therefore the aim of this study was (1) to study the quantitative implication of GPs regarding syphilis and gonorrhoea epidemiology in Flanders; and (2) to examine the opportunities and limits of the Intego database for STI surveillance in this regard.
First, the setting will be explained: Aspects of the health system, as far as they are important for surveillance, and characteristics of the major surveillance systems which include STIs will be addressed, before characteristics of the Flemish Intego general practice database are described. After outlining the study design, epidemiological data on syphilis and gonorrhoea are presented in comparative perspective (Intego versus mandatory notification in Flanders) and discussed with regard to the Belgian health system.

The Setting (I): Aspects of the Belgian Health System and STI Surveillance
In Belgium, -compulsory health insurance is combined with a mostly private system of health care delivery, based on independent medical practice, free choice of physician and predominantly fee-for-service payment‖ ( [5], pxxv). Thus, general practitioners (GPs) do not play a role as gatekeepers to specialist care. Diagnosis and treatment of sexually transmitted infections (STIs) are also provided by gynaecologists, dermatologists, urologists, centres for medical students, family planning centres, STI clinics, and AIDS Reference Centres (ARCs) ( [6], p16). Furthermore, surveillance of infectious diseases is a regional legal competency. Except for the national HIV cohort, surveyed by the Belgian Scientific Institute of Public Health [7], there is no comprehensive The Flemish Intego network at the Department of General Practice of the Catholic University Leuven is -the only operational computerized morbidity registration network in Belgium based on general practice data‖ [16]. Over 90 Intego GPs (see Table 1), evenly spread throughout Flanders, collect data on about 2% -of the Flemish population representative in terms of age and sex‖ [16] (details see Truyers et al. 2014 [16], Truyers et al. 2015 [17], and Vaes et al. 2015 [18]). The research team at the Academic Centre for General Practice of the Catholic University Leuven coordinates the network and analyses the data. All Intego GPs work with the proprietary software programme Medidoc ®. Every year, all Flemish GPs who work with Medidoc are invited to contribute to the Intego network. Of all GPs responding, quality controls are carried out in order to determine which practices belong to the reference group whose data is used for further analyses [19].
Privacy procedures in place imply that GPs sent their data to a Trusted Third Party (TTP) which assigns codes to the patient identifier and the practice from which the data originates ( [20], pp3-4).
Therefore the research team at the Academic Centre for General Practice of the Catholic University Leuven does not know from which patient and from which GP the data originates.
A few practices leave the network every year and need to be replaced by new practices, so that the number of practices and GPs slightly fluctuates, but the total number of GPs remained over 90 throughout the observation period ( [17], p12). From the yearly contact group, the practice population is calculated by using a correction factor for non-attenders, the group which does not visit their general practitioner in a given period. The correction factor is based on reimbursement data from statutory health insurance, provided by the Intermutualistic Data Agency (IMA-AIM) (details see Bartholomeeusen et al. 2005 [21]). The Intego practice population in the observation period 2009-2013 fluctuated around 150,000, while Flanders had between 6.2 and 6.4 million inhabitants (see Table 1). Omnio status (for households with low income) as a socioeconomic variable had not yet been provided during the observation period [19], so that this data is not available for this study.
Intego GPs routinely register all new diagnoses which are collected together with information on the patient from GPs' personal computers and entered into a central database [18]. GPs are requested to encode clinical labels (keywords) offered by the software programme. To each clinical label, Medidoc assigns a programme-specific internal Medidoc code and a -diagnostic group‖ code. Furthermore, it links new diagnoses to the International Classification of Primary Care (ICPC-2) and the International Statistical Classification of Diseases and Related Health Problems, 10th Revision (ICD-10) [18].
Informed consent was not necessary for this type of study. -Intego procedures were approved by the ethical review board of the Medical School of the Catholic University of Leuven (N ML 1723) and by the Belgian Privacy Commission (no SCSZG/13/079)‖ [18].

Study Design, Material and Methods
This retrospective cohort study used Intego data of a 5-year time period from 1 January 2009 to 31 December 2013 from the Intego reference group. The relevant gonorrhoea and syphilis Medidoc codes were selected (see Supplementary Table 1) and cases with these codes included. Two parts of the database were extracted: The first part included the patient numbers, the Medidoc codes, the diagnostic groups gonorrhoea and syphilis, and the start date of the respective diagnosis. The second part included the patient number, the year of birth and sex. The two parts were merged, based on the patient number.
The number of gonorrhoea and syphilis infections was counted as cases (episodes), consisting of one or more patient consultations for the same medical diagnosis. Following the example of Suijkerbuijk and colleagues on chlamydia [22], a second episode with the same diagnosis for the same patient was counted only as a new case after an interval of at least two months after the first diagnosis. We defined the two month-interval as 62 days. Gonorrhoea-syphilis co-infections were defined as registrations of syphilis and gonorrhoea with a beginning date of 7 days or less apart from each other.
One case was excluded due to a software error in which the Medidoc code L09001 for -locomotor congenital syphilis‖ was assigned to -arm pain not further described‖. Six observations with Medidoc code X09131 (-condylomata lata X‖) and 11 observations with Medidoc code Y09131 (-condylomata lata Y‖) were not included. -Condylomata lata‖ describes lesions during secondary syphilis [23]. In Medidoc, the diagnostic group -condylomata‖ includes Medidoc codes for warts of venereal origin and/or ano-genital location including condylomata acuminata and condylomata lata. For these 17 patients, no diagnostic code for syphilis was registered during the observation period. For comparison, the Agency for Care and Health provided data on annual syphilis and gonorrhoea cases obtained by mandatory notification, differentiated by age and sex.

Data Analysis
Descriptive analysis was performed with STATA 12.0 and Excel 2010. For gonorrhoea and syphilis, the number of cases per year was calculated and Poisson confidence intervals (CIs) applied.
For privacy reasons, only the year of birth was given in the Intego database. In order to estimate the age class, July 1 was assigned as fictitious birthday to all patients in order to keep the error of the estimated age within the range of half a year or less. Then the estimated age was calculated by subtracting the fictitious birthday from the beginning date of the diagnosis. Four age classes were chosen for comparison with mandatory notification data.
Taking the age group 45+ years as reference group; odds ratios (ORs) with regard to the other age groups were calculated. For age groups which have fewer than 5 cases, no OR is presented due to the lack of precision.
Due to small numbers obtained, the annual numbers of syphilis and gonorrhoea cases were pooled and averaged for 2009-2013 and the frequency per 100,000 practice population calculated.
For comparison, the annual gonorrhoea and syphilis frequencies in the Flemish population were calculated from the figures of the Agency for Care and Health, Flanders ( [1] and unpublished data), and population statistics taken from the Belgian Federal Public Service Economy [24]. 95% Poisson confidence intervals were applied as well.

Results
During the observation period 2009-2013, 91 gonorrhoea were registered in the Intego database (see Table 1). For one Intego patient, two gonorrhoea cases were registered, one in  Table 1).  Table 1).

Annual Number of Gonorrhoea and Syphilis Cases Registered per 100,000 Population
In the Intego database, the annual number of gonorrhoea cases per 100,000 practice population

Sex and Age Distribution
In both databases, the majority of male gonorrhoea patients were 25 Table 3a-c).

Annual Gonorrhoea and Syphilis Cases Averaged 2009-2013 per 100,000 Population, Flanders, Intego Database and Mandatory Notification
In the Intego database, the average annual gonorrhoea frequency per practice population for the 5-year-period 2009-2013 was 11.9 per 100,000 population, with 19.9 for men and 3.9 for women.  Table 4).

Discussion
This retrospective descriptive study gives new insight into the extent of GP involvement regarding gonorrhoea and syphilis in Flanders as well as the opportunities and limits of the Intego database. Strengths of this study are that the data was collected as part of routine activities as Intego GPs. No additional efforts of GPs were necessary to obtain the data, and there was no special focus on STIs, thus reducing possible bias.

Verbrugge and colleagues interpret the annually reported syphilis and gonorrhoea cases in the
Sentinel Network of Microbiological Laboratories per reference population as reported incidence [6,9].
We also interpret annual Intego and mandatory notification figures per year and reference population as reported incidence, since new cases came to the knowledge of and were registered by Intego GPs and the Agency for Care and Health, respectively, whether or not these cases constituted new infections or newly discovered but previously acquired ones.
The interpretation of the data requires caution. The study situates itself within the context of incomplete figures concerning Belgian STI surveillance and a substantial variability between different European countries [25]. The data confirm the results of Laisnez et al., who found that the GP in East and West Flanders treated gonorrhoea in the large majority of cases [3,4], as well as the dilemma that most gonorrhoea infections are seen by the GP but that it is a rare disease for the GP and a challenge to keep him/her up-to-date with the latest relevant guidelines [3]. in that year ([6], p47 to the knowledge to the GP, meaning he/she was involved either based on his/her own diagnosis or on information from other healthcare providers.

Challenge of Transferability to all Flanders
While the practice population is representative for the Flemish population, we do not know whether STI investigation strategies and communication patterns with other healthcare providers of Intego GPs are representative for Flemish GPs as well. It is likely that Intego GPs are more than average committed to their profession. If their STI detection strategies are better than average, then the annual Intego figures per 100,000 reference population might be higher than those obtained from GPs in all Flanders.

Small Annual Numbers
The small number of cases registered annually inevitably led to wide confidence intervals, In their study on cancer registries, Takiar and colleagues argued -that when the population of the registry is around 150,000 then the incidence rate of even 10 should be viewed with reservations‖ ( [33], p660). The patient population of the Intego database has about this order of magnitude.

Opportunities for Further Research
Further research could look into ways to improve the cooperation between the GP and the Simple adjustments to the collection of Intego data could lead to further insights. (1) The possibility to mark a diagnosis as -reported‖ by another healthcare provider would show the extent of other health care providers contributing to diagnoses in general practice. Furthermore, reporting culture of specialists could be examined. (2) Assigning the criterion of a period of more than two months between two registrations of the same diagnosis in order to distinguish one episode of care or one case from another is necessarily arbitrary. The -Weekly Returns Service‖, today under the name Royal College of General Practice Research and Surveillance Centre (RCGP-RSC) [35], a general practice network serving England, counts episodes as -first‖ (first diagnosis ever), -new‖ (new episode of care) and -ongoing‖ (another registration during an episode of care having started previously) ( [36], p88). Taking over these categories could add precision to case definitions in the network.

Conclusion
The Intego database offers epidemiological data on selected STIs from routine registration in general practice in Flanders, giving an indication on quantitative GP involvement in STI control.
Although reliability and precision are limited by small numbers, especially for syphilis, comparison with mandatory notification suggests that GPs came to know the large majority of gonorrhoea episodes, but not the majority of syphilis episodes, for which reporting from specialist care seems not very developed. Opportunities for further development of the Intego database include discriminating GP-diagnosed episodes from episodes registered there but diagnosed by other healthcare providers.

Funding/Declaration of Interest Statement
Intego is funded on a regular basis by the Flemish Government (Ministry of Health and Welfare