GPs’ awareness of pregnancy: trends and association with hazardous medication use

Summary

Among 140 976 selected pregnancies, for only 48% a pregnancy confirmation was recorded in the GP records, indicating GPs’ (potential) awareness of pregnancy. A statistically significant increase was observed from 28% in 2004 to 63% in 2020. The large majority (78%) of these confirmations happened during the first trimester. In 3% of all included pregnancies, the GP prescribed highly hazardous medication with teratogenic effects that should have (temporarily) been avoided, with pregnancy being GP confirmed for only 13% at the first occurrence of such a prescription. In 11% of patients with a highly hazardous prescription before pregnancy confirmation this prescription was repeated after confirmation. Comparative analyses showed that women without a GP-recorded pregnancy confirmation were 59% more likely to be prescribed this highly hazardous medication. Even though this study demonstrated that the absence of a recorded pregnancy contraindication in the GP system for automatic drug surveillance significantly increased the odds by 41% of prescribing highly hazardous medication, such an active link was created in only 13% of pregnancies (increasing from 0% to 22% during the study period). When using all available coded and uncoded electronic records from the GP information system, a pregnancy indicator was available for 70% of pregnancies. However, in 2020, 23% of pregnancies were still not registered in the GP information system. Only 1% had a GP consultation specifically set up for preconception counselling recorded in the 12 months preceding pregnancy.

Strengths and limitations

A major strength of this study was the use of over 15 years of routinely collected data from a unique and large population- based linked cohort, shown to be representative of the Dutch population.¹⁷ The timing of registered records relative to pregnancy could be accurately assessed based on information from the different databases. This study thereby provides a unique, contemporary perspective on GP involvement in pregnancy in real-world clinical practice.

There are, however, several limitations. The qualitative concept of GPs’ awareness was quantified by using available electronic patient records from the women’s medical file, however, this should be regarded as ‘potential’ awareness. GPs may have been aware of pregnancy, but did not record this for a variety of reasons (for example, time constraints, deemed redundant, system difficulties, or because no letter was even received from the midwife). It has been acknowledged, however, that GPs rely on the information registered in their systems for providing accurate patient care, for example, in case of transfers to other caregivers.^27,28

A common challenge in using administrative data is defining drug exposure or compliance. Prescription records can only approximate actual exposure and, particularly during pregnancy, prescriptions may not be filled or drugs may be discontinued. Although use of hazardous medication could therefore have been overestimated, it is not expected to have altered the conclusions in relation to GPs’ awareness, since sensitivity analyses using medication that was dispensed in the pharmacy provided similar results. Underestimated drug exposure is likely because specialist-prescribed and over-the- counter drugs were not included, however, the intended focus of the current study was on GPs’ prescription practices. Confounding by indication could not be ruled out in the definition of hazardous medication. Although pregnancy-driven prescriptions were specifically excluded where possible, some of the hazardous medication may still have been prescribed because of pregnancy. For example, as observed in the drug-level assessment, misoprostol may have been prescribed to induce abortion. However, since prescribing for these reasons would normally occur in secondary care, they are assumed to be prescribed for other indications in most cases. Unfortunately, data on the indication of use was only available for a small proportion of prescriptions. Absolute rates presented as part of the drug-level assessment should therefore be interpreted with caution and conclusions can be drawn from relative comparisons by GPs’ awareness status.

Comparison with existing literature

This study contributes valuable new evidence to the role of GPs in daily clinical practice during pregnancy. Although existing literature on the outcomes of interest is scarce, one previous Dutch study reported a recorded diagnosis for pregnancy in the GP records for 41% of births from 2007–2009, which is very similar to what is observed in the current study.³ Taking into account slight differences in study period and design, the findings of prescribed (highly) hazardous medication were in agreement with previous studies.^29–32

To the authors’ knowledge, this is the first study to assess the association between GPs’ awareness of pregnancy and prescribing hazardous medication. Although no comparison information is available on the registration of pregnancy as a contraindication in the GP information system, efforts have been made to measure the quality of GP registrations by defining a set of quality indicators, such as the proportion with a recorded contraindication. This was reported to vary among GP practices and thus they were instructed to critically review their daily habits for registering contraindications.³³ More generally, maintaining medical records has been acknowledged as a fundamental part of a doctor’s duties in providing patient care and despite this importance, it is often given low priority.^27,28

Similar to this study, many other previous studies have concluded that delivery of preconception care is inadequate.^{11,12,34–36} A Dutch survey conducted among GPs and midwives reported that only 0.7% of GPs systematically invited patients for a formal preconception care consultation.¹¹ Although this is similar to the 1% observed in the current study, 20% of those GPs who were surveyed¹¹ indicated that they performed preconception care in a standardised manner, which is probably not captured by the strict definitions used for a preconceptional GP consultation in the current study. For example, it is likely that a GP may counsel women preconceptionally as part of a consultation (coded) for something else such as the underlying condition. Although Dutch guidelines have clearly advocated standardised preconception care for some time,³⁴ collaboration between GPs and other caregivers is advised. There seems to be a shift towards a more public, programmatical approach incorporated in the daily care provided that may explain the low occurrence of systematic preconceptional GP consultations observed in the current study.^37,38

In comparison with other countries with similar healthcare systems, including the UK, similar conclusions have been drawn on the need for shared, multidisciplinary pregnancy-related care programmes in which preconception care should be offered.^10,13,14,39

Implications for research and practice

The finding that women without a GP-recorded pregnancy confirmation were significantly more likely to be prescribed hazardous medication indicates a potential awareness issue at the time these drugs are prescribed, placing women and their babies at avoidable risk of exposure to teratogens. Although pregnancy registration by GPs improved over the years, inadequate use still seems to be made of the available information systems for appropriate drug surveillance. The key challenge for improved registration lies with the shared responsibility, in which collaborative care is pivotal. The authors pose three main implications based on the study findings in combination with the existing evidence.

First, caregivers should be supported and educated in maintaining accurate and readily available patient records for effective communication and information transfer to other involved caregivers. Specifically in pregnancy this requires continuity of care by documenting medical records on a daily basis to prevent use of harmful medication because of delayed or incomplete record keeping.

To achieve this, the second implication relates to the electronic information systems used to maintain records. There should be clear and standardised procedures for recording and communicating information so that healthcare providers know what is expected. The differences in GPs’ awareness observed between GP information systems suggest the need for further standardisation of systems. The increased availability of pregnancy indicators when using all available records from the GP information system implies difficulties in choosing the appropriate GP reference tables for registering pregnancy, obstructing GPs’ awareness.

Further simplification would be helpful, for example, by automatically establishing an active contraindication in that specific GP reference table in case of pregnancy confirmation, blocking the prescription of certain high-risk drugs and avoiding alert fatigue among caregivers. A financial incentive was provided by the Dutch government in 2012 and 2013 for improvement of coded registration in GP practices,⁴⁰ which is also reflected in the increased coded pregnancy indicators in the second half of the study period. During the study period the conversion of GP records from handwritten to computerised also took place.

Third, public awareness about the potential risks of medication used during pregnancy should be improved by means of population-wide education incorporating collaborative preconception care. In addition to caregivers acknowledging their duty here, this would ultimately increase women’s self-awareness recognising their own responsibility in timely informing caregivers about (planned or unplanned) pregnancy, so that appropriate action can be taken. When prescribing hazardous medication, raised awareness would make prescribers more actively enquire about pregnancy, even in the case of repeat prescriptions. Interventions should be set up in such a way that women are informed about the potential pregnancy risks of the medicines they use as early as possible, so that the patient is alert if she is considering conceiving. Whether interventions have the intended positive effects should be evaluated according to predefined targets. Pregnancy prevention programmes for highly hazardous drugs should be continuously evaluated and set up as needed.⁴¹

Future qualitative research among GPs, midwives, and pharmacists would be very useful to further estimate the scale of the posed awareness issue and associated aspects, such as the women’s lack of awareness of pregnancy, shortcomings of information systems, and the barriers perceived in collaborative care.