Abstract
Objective To understand the possible association between media coverage and changes in the dispensation of doxylamine-pyridoxine in Canada.
Design Cross-sectional time-series analysis using data from the IQVIA CompuScript database.
Setting Ten Canadian provinces.
Participants Family physicians, general practitioners, and obstetrician-gynecologists.
Main outcome measures Data on the estimated total volume of doxylamine-pyridoxine prescriptions dispensed by retail pharmacists in Ontario and across Canada between July 2016 and May 2022 were used for a time-series analysis. Birth data obtained from Statistics Canada were used to adjust for pregnancy rates. Finally, the possible impact of media coverage in January 2018 on doxylamine-pyridoxine use was assessed with interventional autoregressive integrated moving average modelling, and dispensation rates of doxylamine-pyridoxine were reported overall and by prescriber specialty.
Results Doxylamine-pyridoxine dispensation decreased by 1.2% (P=.015) overall in Ontario but did not change significantly (P=.064) across the rest of Canada. Out of 619,720 total doxylamine-pyridoxine prescriptions dispensed, 391,722 (63.2%) prescriptions were written by family physicians or general practitioners in Ontario, and there was a decline of 2.4% (P=.010) in the dispensation of doxylamine-pyridoxine based on prescriptions written by this group that lasted for 6 months before returning to previous dispensation rates. There was no significant change in the dispensation of doxylamine-pyridoxine based on prescriptions written by obstetrician-gynecologists (decline of 0.5%, P=.235) in Ontario.
Conclusion Despite nationwide media attention questioning the efficacy of doxylamine-pyridoxine in early 2018, trends in dispensation rates of doxylamine-pyridoxine were only modestly affected and only in Ontario. Within Ontario, a small but significant reduction in doxylamine-pyridoxine dispensations was noted based on prescriptions from family physicians or general practitioners but not based on those written by obstetrician-gynecologists. Factors associated with the ongoing use of doxylamine-pyridoxine despite evidence of lack of benefit require further research.
Doxylamine-pyridoxine is commonly used to treat patients experiencing nausea and vomiting of pregnancy (NVP), otherwise known as morning sickness. Although NVP varies in duration and severity, it affects 70% to 80% of pregnant patients.1 While most patients with NVP experience symptoms within the first trimester, some experience prolonged symptoms.1 However, the exact cause of NVP is unknown and finding an optimal treatment can be difficult.1 Symptoms of NVP generally occur within the first few weeks of pregnancy, ranging in severity from mildly uncomfortable to severe NVP, known as hyperemesis gravidarum.2 Despite varying severity of NVP, studies have shown a decreased quality of life among patients experiencing NVP symptoms in addition to increased hospital admissions, health care visits, and prescribed medications.3 The first-line recommendation for treatment of patients with mild NVP symptoms is nonpharmacologic methods, namely behavioural and dietary modifications such as avoiding triggers, eating smaller meals, and resting.3,4 For moderate to severe symptoms, pharmacologic treatment is recommended.3 Pyridoxine monotherapy (a form of vitamin B6) or doxylamine-pyridoxine combination therapy (doxylamine is an antihistamine) is recommended as first-line pharmacologic treatment for patients with NVP owing to their safety and efficacy profiles.5 Metoclopramide, an upper gastrointestinal motility stimulant, may be safe to use but evidence for efficacy is limited.5
Doxylamine-pyridoxine has been used in Canada since 1957 and is available as a fixed-dose combination product of 10 mg of doxylamine and 10 mg of pyridoxine for treatment of NVP.6 Doxylamine-pyridoxine tends to be prescribed by family physicians and general practitioners (FPGPs) and by obstetrician-gynecologists (OBGYNs). Obstetrician-gynecologists specialize in caring for pregnant patients and as a result are more likely to see patients with higher-risk pregnancies, including those presenting with severe NVP.7 In comparison, FPGPs tend to provide comprehensive care for families, which may include pregnancy care and delivery.7 However, FPGPs are more likely to see patients with low-risk pregnancies without multiple comorbidities or severe NVP.7 In Ontario first-trimester pregnancy care is provided by FPGPs or walk-in clinics, resulting in a lower frequency of OBGYN visits in the beginning of pregnancy when NVP tends to occur.8
Within 5 years of the US Food and Drug Administration’s 2013 approval of a specific formulation of doxylamine-pyridoxine, a group of researchers (led by N.P.) based in Toronto, Ont, used original data from Health Canada to replicate the analysis reporting superiority of the drug versus placebo.2,6 The re-analysis was conducted due to discrepancies between the prespecified plans for the trial and the published report.2 The re-analysis employed the original statistical analysis plan and individual patient-level data.2 Although the re-analysis found a trend toward reductions in NVP symptoms with doxylamine-pyridoxine versus placebo, the magnitude of the difference was substantially smaller than the prespecified clinically important difference.2 The observed difference was either 0.73 or 0.38 on the 13-point Pregnancy-Unique Quantification of Emesis scoring system, depending on the handling of missing data, while the clinically important difference appeared to have been set at 3 points.2 Prior to the re-analysis being published, Health Canada had conducted a review of doxylamine-pyridoxine in 2016 that found the original study results were not definitive and the clinical benefit was unclear.6
Research re-examining doxylamine-pyridoxine was covered by several news platforms in January 2018 including the Globe and Mail, CBC News, and the Toronto Star.9-11 The Hospital for Sick Children in Toronto released a public disclosure addressing concerns raised about the researcher who worked at the hospital and contributed to the original published trials on doxylamine-pyridoxine.12 The news release addressed questions regarding potential conflicts of interest based on the researcher’s association with the manufacturer of the combination drug.12 However, regulatory bodies did not address this update in any sort of published guidance.
Public communications published through various platforms and media sources have been identified as being associated with shifts in medication prescribing trends and health service usage.13-18 Media reports can play a crucial role in bridging communication between regulatory agencies, manufacturers, and patients—potentially affecting patient and prescriber belief systems and thus potentially altering prescribing patterns. The impact of public communications may vary according to factors such as the delivery and verbiage of the communication.17 For example, a study on antipsychotic prescribing in Ontario nursing homes following a series of newspaper articles published on this topic in 2014 noted a sudden decrease in prescribing within 6 weeks of the series being published; the percentage of nursing home residents with antipsychotic medication prescriptions continued to decline steadily following the media reports through to the end of the study in March 2016, with a net reduction in antipsychotic prescribing of 6.0%.13 Similarly, statin discontinuation in Denmark increased substantially following negative media attention in a Danish newspaper.14
Given widespread media attention on doxylamine-pyridoxine and the need to communicate information about drug safety and effectiveness to the public, it is important to understand whether news reports were associated with changes in doxylamine-pyridoxine use. Our objective was to assess dispensation rates of doxylamine-pyridoxine before and after the media reports were published in Canada9-11 and specifically in Ontario according to prescriber specialty.
METHODS
Study design
We conducted a time-series analysis of monthly purchasing data for doxylamine-pyridoxine within Ontario and across Canada between July 2016 and May 2022 using IQVIA’s Canadian CompuScript database. Time-series analyses are commonly used to assess the impact of health interventions over long time spans. Data are repeatedly measured to account for underlying trends and to identify variation in the data before and after an intervention. We extracted the estimated total number of prescriptions dispensed across retail pharmacies to assess doxylamine-pyridoxine dispensing trends in Ontario and across Canada before and after widespread media coverage9-11 in January 2018.
Data sources
IQVIA enables data access to support evidence-based research regarding drug development, patient outcomes, and product safety. IQVIA’s Canadian CompuScript database covers drug data across all Canadian provinces, providing information on the estimated total number of prescriptions dispensed by retail pharmacies, the number of units (eg, tablets, capsules), and the sum of all costs of prescriptions dispensed. We obtained data on the total volume of doxylamine-pyridoxine purchases sold by retail pharmacists in Ontario and across Canada. We leveraged birth data obtained from Statistics Canada to adjust for pregnancy rates; that is, we obtained the number of live births as an indirect measure of pregnancies and adjusted the dispensing rate based on the number of live births.
Analysis
We reported data on monthly doxylamine-pyridoxine dispensations across Canada and within Ontario, which we further stratified by prescriber specialty (FPGPs and OBGYNs). We used interventional autoregressive integrated moving average (ARIMA) models to determine the statistical significance of changes in doxylamine-pyridoxine dispensations. Autoregressive integrated moving average models are often used to understand past trends or predict future trends based on a measure at different points in time. These models can account for underlying trends, autocorrelation, and seasonality.19 They are increasingly used to examine the impact of population-level health interventions in time-series analyses.19 In this case we used an ARIMA model to assess variation in dispensation trends following media uptake surrounding doxylamine-pyridoxine based on past trends in purchasing data. We used a step function to represent the media coverage of doxylamine-pyridoxine in January 2018. Data analysis for ARIMA modelling was conducted using SAS software, version 9.4.
RESULTS
Monthly doxylamine-pyridoxine dispensation rates in Canada
Overall, the monthly average number of doxylamine-pyridoxine prescriptions dispensed at Canadian retail pharmacists was 18,523. Of those prescriptions, 14,636 (79.0%) were written by FPGPs and 3887 (21.0%) were written by OBGYNs. Dispensation rates of doxylamine-pyridoxine did not change significantly across Canada following the January 2018 media coverage (an increase of 1.67%, P=.064, Figure 1)2,9-11 or across prescriber specialties (an increase of 0.48% [P=.612] among OBGYNs and a decrease of 2.4% [P=.099] among FPGPs).
Estimated monthly dispensation rates of doxylamine-pyridoxine in Canada between July 2016 and May 2022
Monthly doxylamine-pyridoxine dispensation rates in Ontario
Between January 2018 and May 2022, the dispensation rate of doxylamine-pyridoxine decreased significantly by 1.2% (P=.015) overall across Ontario (Figure 2).2,9-11 Of 619,720 doxylamine-pyridoxine prescriptions dispensed in Ontario, 391,722 (63.2%) were written by FPGPs and there was a significant decline of 2.4% (P=.010) in doxylamine-pyridoxine dispensations based on prescriptions written by this group. Approximately 103,472 dispensations were based on prescriptions written by OBGYNs in Ontario, accounting for 16.7% of total doxylamine-pyridoxine dispensations. There was a nonsignificant decline of 0.5% (P=.235) in dispensations based on prescriptions from OBGYNs in Ontario (Figure 3) following media coverage.2,9-11
Estimated monthly dispensation rates of doxylamine-pyridoxine in Ontario between July 2016 and May 2022
Estimated monthly dispensation rates of doxylamine-pyridoxine in Ontario between July 2016 and May 2022, stratified by prescriber specialty
DISCUSSION
We observed modest but statistically significant changes in doxylamine-pyridoxine dispensation in Ontario after media reports on research re-examining use of the combined medication. However, doxylamine-pyridoxine dispensation rates did not change across Canada. As monthly fluctuations in dispensations appeared to have been greater across Canada than in Ontario (Figures 1 and 2),2,9-11 shifts may have been easier to detect in Ontario.
Previous analyses examining the impact of safety communications on medication prescribing indicate the impact of public communications about drug advisories may vary based on how information is communicated.13-18 An interrupted time-series analysis evaluating the impact of drug advisories on drug utilization in Canada, Denmark, the United Kingdom, and the United States published in 2022 found the impact of advisories to be modest and variable.17 Variability of the impact may depend on the content of the communication, severity of risks, identification of patients at risk, and strength of evidence.17 Communications regarding doxylamine-pyridoxine questioned the effectiveness of the medication, emphasizing the review2 published in January 2018. Given that the research investigating the efficacy of doxylamine-pyridoxine was based in Toronto it is possible the findings may have received greater attention within the province.
What might explain why dispensation rates of doxylamine-pyridoxine did not change significantly across Canada, despite reports pointing to lack of efficacy? One explanation might be the lack of other therapies approved for NVP in Canada, in terms of being able to provide patients with pharmacologic treatment options. Patients with NVP are typically advised to rest, stay hydrated, avoid triggers, and eat smaller meals.3
The variation we noted between FPGPs and OBGYNs in Ontario, with a significant reduction in doxylamine-pyridoxine dispensations based on prescriptions from FPGPs but not OBGYNs, may point to potential differences in demographic characteristics of patients they treat. For example, if OBGYNs happen to be seeing patients with more severe NVP symptoms, those patients might be more inclined to try or keep using doxylamine-pyridoxine to address their symptoms, even with questions about its efficacy, and thus we might expect to see less of a shift in the volume of doxylamine-pyridoxine dispensations based on prescriptions from OBGYNs. Also, as the 2018 review of doxylamine-pyridoxine2 was led by a family physician (N.P.), that research may have been more widely communicated to or read by colleagues in the family medicine community versus other specialists.
Limitations
This study examines a unique landscape following widespread media attention surrounding doxylamine-pyridoxine use across Canada and within Ontario. However, several limitations of this study arose from the data. First, the Statistics Canada birth data did not include data from Nunavut, Yukon, and the Northwest Territories. Although this excludes only a small portion of the Canadian population (less than 1%),20 we are unable to generalize our findings to people living in those 3 territories. Furthermore, due to the nature of our data we are unable to identify differences in health care practices and clinic access between provinces. Without understanding key differences between provinces and potential variation in prescribing practices, we cannot identify reasons for why dispensation of doxylamine-pyridoxine did not shift significantly in provinces other than Ontario. Future investigation of prescribing practices and health care system differences between provinces would add notable insight into this work. Additionally, IQVIA Canadian CompuScript excludes information regarding patient and prescriber demographic characteristics and clinical data, so we were unable to identify whether shifts in doxylamine-pyridoxine dispensation rates may have varied based on factors such as prescriber sex, patient or prescriber age, patient comorbidities, and low- or high-risk pregnancy status. Similarly, we were not able to identify whether dispensation trends may have been affected by potential behavioural changes, such as delays in seeking or accessing care during the COVID-19 pandemic.21 Further exploration of such behaviour and patient and prescriber demographic characteristics to identify whether certain subgroups are more or less likely to have shifts in prescribing and filling of prescriptions would benefit this work. Finally, due to the observational nature of this research, we cannot make causal inferences from these results.
Conclusion
Following nationwide media attention questioning the efficacy of doxylamine-pyridoxine, there was a small but significant negative shift in dispensation trends for this medication in Ontario, but not in other parts of Canada. Within Ontario, a small but significant reduction in doxylamine-pyridoxine dispensations was noted based on prescriptions from FPGPs but not for those written by OBGYNs. Further exploration of patient and prescriber characteristics and behaviour may help explain these shifts and shed light on factors associated with the ongoing use of doxylamine-pyridoxine despite evidence questioning its benefits.
Acknowledgment
This study was supported by funding obtained from the Canadian Institutes of Health Research. The analyses, conclusions, opinions, and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources. The statements, findings, conclusions, views, and opinions contained and expressed in this research article are based in part on data obtained under licence from the following IQVIA Solutions Canada Inc information services: CompuScript database. All rights reserved. The statements, findings, conclusions, views, and opinions contained and expressed herein are not necessarily those of IQVIA Solutions Canada Inc or of any of its affiliated or subsidiary entities.
Footnotes
Contributors
Cherry Chu and Setayesh Yazdani conducted the analysis. Shenthuraan Tharmarajah created figures and tables. Shanzeh Chaudhry drafted the manuscript. Dr Mina Tadrous supervised the project, provided direction, and gave feedback. All authors reviewed and approved the final manuscript.
Competing interests
None declared
This article has been peer reviewed.
Cet article a fait l’objet d’une révision par des pairs.
- Copyright © 2025 the College of Family Physicians of Canada
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.