Introduction

Chlamydia trachomatis (chlamydia) is one of the most reported bacterial sexually transmitted infections (STI) worldwide [1]. Chlamydia infections can lead to complications such as pelvic inflammatory disease (PID), which subsequently could lead to tubal factor infertility (TFI) or ectopic pregnancy (EP) [2]. Globally, the age-standardized rate of PID prevalence was 53.19 per 100,000 population in 2019 and for EP it was 342.22 per 100,000 population [3]. A study in the United Kingdom estimates that every 1,000 chlamydia infections in women aged 16–44 years, on average, results in 171 extra episodes of PID, 2 more EP and 5.1 more women with TFI at age 44 years [4].

To timely detect chlamydia infections to prevent complications, many countries implemented chlamydia trachomatis (chlamydia) control activities ranging from active case finding to screening programmes [5–7]. However, there is an ongoing debate about the impact of these ‘test and treat’ strategies on reducing chlamydia population prevalence and complications including PID, EP and TFI [7–10]. A possible way to increase knowledge about the impact of control strategies on occurrence of complications is by examining population-level associations between chlamydia control activities and clinical complications [11, 12]. Several countries compared the association between control activities and clinical complications [11, 12]. The authors stress the urgent investment needed to establish surveillance systems to record and monitor trends of chlamydia related complications, such as pelvic inflammatory disease (PID) over time [11]. To our knowledge, no country has implemented routine surveillance of chlamydia related complications.

In the Netherlands, there is a debate about the public health relevance of active case finding by testing for asymptomatic chlamydia infections [7, 10]. This debate is based on lack of practice-based evidence that widespread testing of asymptomatic infected individuals can reduce population prevalence [8, 9]. Also, a number of potential harms from overdiagnosis and overtreatment have been reported, such as psychological stress and potentially unnecessary use of antibiotics, which could lead to antimicrobial resistance [7, 10]. As a result, van Bergen et al. [10] suggest that future strategies should reduce, rather than expand, widespread testing for asymptomatic chlamydia infections. Some countries have already implemented small changes to their testing guidelines, e.g. focussing more on reducing harms from untreated chlamydia infection rather than onward transmission [13]. Australia replaced its focus from testing uptake to strengthening management of diagnosed infections, which involves improving retesting, partner management and PID diagnosis [14].

In the Netherlands, guidelines recommend chlamydia testing for persons experiencing symptoms, their contacts, and risk groups regardless of symptoms such as persons who had unprotected sex [15]. But in the near future, guidelines will recommend to limit asymptomatic chlamydia testing for both men and women. Such a change in chlamydia control activities might impact chlamydia surveillance, which is based on people getting tested for chlamydia at Sexual Health Centres (SHC) or the general practitioner (GP) [16]. This highlights the need for setting up alternative ways to monitor trends in chlamydia infections. In addition, better insight into chlamydia complication trends is needed when changing the focus from infection control to disease management [10]. Currently, no national surveillance system for PID, EP and/or TFI is in place.

We aimed to investigate the possibilities to introduce long term chlamydia complications (e.g. PID, EP and TFI) under surveillance in the Netherlands according to relevant stakeholders that diagnose these complications. More specific, we aimed to understand the process of diagnosis and registration of these complications and investigate the view of stakeholders on the relevance of a complication surveillance system when changes of chlamydia testing are in place.

Methods

Results

Fifteen in-depth interviews with a purposive sample of gynaecologists (n = 3), medical professionals under supervision of a gynaecologist (n = 2: in vitro fertilization (IVF)-doctor and fertility doctor), GPs (n = 6), doctors at a regional SHC (n = 3) and emergency room (ER) doctor (n = 1) were conducted. Gynaecologists, IVF and fertility doctors were grouped and referred to as gynaecologists (n = 5). Ten respondents were female (67%), and ten respondents were 20–39 years old, five were 40 years or older. Respondents worked in seven of twelve provinces in the Netherlands.

Pelvic inflammatory disease (PID)

Diagnosis.

According to all respondents, patients with symptoms related to PID most often consult a GP and/or gynaecologist, sometimes via the emergency room. All GPs see patients with suspected PID, and all but one GP would always refer a patient to the gynaecologist for further diagnosis and treatment (Table 1). This GP reported occasionally diagnosing and treating PID without interference of a gynaecologist. All respondents acknowledge the difficulty of diagnosing PID, as symptoms are often non-specific and there is no easy, non-invasive diagnostic test. “it is not as if there is an easy test that tells you it is a PID” (Gynaecologist).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Overview of diagnosis and referral of complications according to medical professionals (n = 15).

https://doi.org/10.1371/journal.pone.0305279.t001

All GPs were aware of the guideline on diagnosing PID. Some GPs strictly adhered to the guideline with respect to PID symptoms, and expect other GPs to diagnose accordingly. “I hardly ever diagnose or register a PID, because I, as I said, almost never see those classical three; stomach ache, fever and an STI” (General practitioner). Other GPs were less conclusive on what specific symptoms to include for a final diagnosis, and talked more about the overall feeling and the patient’s story. Most respondents suspect a PID based on symptoms and anamnesis (i.e. GPs and gynaecologists) and, if PID is suspected, perform a chlamydia PCR test to identify a current chlamydia infection. GPs were not unanimous on always performing a chlamydia PCR test, one GP assumed that gynaecologists test for a chlamydia infection after the GP referred the patient to them and another GP said they will only test if the patient has a high STI risk. All SHC doctors reported seeing an occasional PID and referring this patient to the gynaecologist for further diagnosis and treatment.

Discussion

According to our respondents, gynaecologists diagnose PID, EP and TFI most frequently compared to other health professionals. Other professions rarely diagnose these complications and mostly refer patients with a suspected complication to the gynaecologist. Emergency doctors only diagnose EP. Most respondents reported unique registration codes for PID and EP, but the coding for TFI is more ambiguous. These are mostly general complication codes, not necessarily related to a (previous) chlamydia infection. Respondents reflected that diagnosis and registration of PID, EP and TFI are handled differently within their profession. Most respondents acknowledged registration in diagnostic codes as a useful surveillance tool.

Our study is the first to explore possibilities for setting up surveillance of PID, EP and TFI by including multiple relevant stakeholders from different regions of the Netherlands. We included both newly started and more experienced professionals with broad coverage in the country. A possible limitation of this study is that doctors who responded were more interested in the theme and might be more knowledgeable about the topics. We do not believe that this influences the results about diagnosing or registration, but they might voice a more positive tone towards setting up a surveillance system. Second, a possible limitation is the sample size. It was difficult to include gynaecologists as they were often occupied with clinical work and were not able to clear their schedule. For this reason, we included professionals in training or under supervision of a gynaecologist. We believe, however, that they have enough experience to adequately answer our questions, as their answers were in line with the three gynaecologists that were included. Based on the interview with the ER doctor, the interviews with other stakeholders, and informal communication, we decided not to include more ER doctors. We concluded that they were not relevant stakeholders in the diagnosis and registration of complications in the Netherlands. Saturation for the main themes was reached after two interviews with SHC doctors, mainly since they explained that they hardly ever diagnose chlamydia related complications. A third interview was conducted to confirm their responses regarding the main themes. However, since network and snowball sampling could result in a more limited specificity, and thus requiring more participants to reach saturation, it might still be possible that other views among SHC doctors exist [21]. Furthermore, we believed that saturation for GPs and gynaecologists was reached for the main themes. For example, professionals were unanimous that diagnosis of complications would not change if chlamydia testing strategies would change (main theme 1). For other topics, such as whether or not all GPs would refer a suspected PID to the gynaecologist, a larger sample size could yield a wider range of views than reported in this article.

Currently two potential existing systems based on registration codes could be implemented for surveillance of complications purposes. First, a sentinel surveillance system, called Nivel (the Netherlands Institute for Health Services Research) sentinel stations, exists in a small proportion (around 8%) of GP offices in the Netherlands with a patient population registered at these practices being representative for the Dutch population in terms of age and gender. This source includes routinely extracted data from electronic health records (consultations and morbidity in ICPC codes) and is already in use to estimate the number of chlamydia diagnoses performed at GPs. Second, for data of patient’s medical records from hospital registries, the Dutch Hospital Data (DHD) can be contacted for a data request for research or surveillance purposes. The DHD uses ICD-10 coding for medical data (e.g. diagnosis and treatment) and has nationwide coverage. All gynaecologists in the Netherlands are linked to a hospital and thus their patient’s medical records are present in the DHD. In our study all hospital staff talked about DBC coding as the standard practice for registering diagnoses. DBC coding is a clustering of ICD-10 codes with an automated connection to the ICD-10 codes used in the DHD. ICD-10 and DBC coding are used by all hospitals in the Netherlands as the standard registration practice for medical records and are a Dutch coding system used for billing healthcare services provided to the patient.

These hospital and GP data sources could be used to implement a surveillance system on long term complications of chlamydia. However, professionals identified potential concerns regarding data quality, as codes for EP and TFI can include other diagnoses and not all PID is registered correctly by the physician in the correct registration codes (e.g. in patient files at GP offices). Furthermore, we also identified concerns regarding representativeness of the registration system, as we observed differences in interpretation of diagnostic criteria within and between professions. For example, one GP was very clear about only diagnosing and registering a PID if a patient presented the three classical criteria (i.e. stomach ache, fever and an STI), whereas another GP would diagnose a PID based on the overall feeling and the patient’s story.

When considering the different perspectives and professions in the diagnosis of complications, we see that most complications are diagnosed by the gynaecologists. Gynaecologists, SHC doctors and the ER doctor think that the GP has the most complete patient files related to complications as they will get feedback letters from other professionals to add to their patient’s file. However, GPs did not feel confident that their patient files are most complete related to PID, EP and TFI compared to the other professions. GPs stated that the registration code for the complication might not always be administered or changed in the patient files after receiving the feedback letter from the gynaecologist. Based on what respondents told us, we feel that gynaecologists, more than GPs, actually diagnose the complications, and properly register them in their hospital databases. Furthermore, as GPs hardly ever diagnose chlamydia complications themselves, but rather irregularly register the diagnosis retrospectively when receiving a feedback letter from the gynaecologist, we expect great overlap between the diagnoses in GP and DHD registration systems. We therefore expect the patient files of gynaecologists to be most complete for monitoring trends of diagnosed chlamydia related long term complications and thus the DHD registration most suitable for complication surveillance.

When considering complication surveillance, an important aspect should be taken into account. Possible observed changes in PID, EP and/or TFI trends cannot be directly linked to changes in incidence of chlamydia infections, as there could be many other explanations for the observed trends. Especially since other infections can also lead to PID, EP and TFI, such as gonorrhoea [22]. Other explanations could include changes in access to health care, (registration of) diagnosis, or changes in treatment of chlamydia infections [23]. Therefore, one should be careful in interpreting trends of PID, EP and TFI when complication surveillance is in place.

Conclusion

Patient files of gynaecologists are likely to be most complete for monitoring trends of diagnosed chlamydia related long term complications in the Netherlands. However, when establishing a chlamydia complication surveillance system, professionals should be engaged in further standardizing diagnosis and registration practices. This will improve the quality and interpretability of complication surveillance and facilitate comparison between countries.

Supporting information

Acknowledgments

The authors would like to thank all respondents for their interesting and insightful thoughts to help our research.