The SafeStart study
This study was a part of the SafeStart interventional trial [21, 22]. Norwegian-speaking, pregnant women in their first trimester were eligible for participation. The SafeStart interventional trial included a total of 229 women who responded to Q1 and Q2 and were included in the analyzes for this study. Of 229 women, 103 were allocated to the intervention group and 126 to the control group. Post hoc power analyzes showed that a sample size of 229 was sufficient to detect a 19% difference in medication use with 80% power. The SafeStart study was conducted according to the CONSORT guidelines [23].
Recruitment
The SafeStart Interventional trial recruited pregnant women between February 2018 and February 2019 through social media, e.g., Facebook, pregnancy-related webpages/forums, and flyers in pharmacies throughout Norway.
Allocation
All women who consented to participate in the SafeStart study were assigned to either the intervention group or the control group (1:1) by a software developed specifically for this project. The software automatically handled the womens` enrollment, group allocation, and distribution of informational emails and online questionnaires.
The intervention group
The women in the intervention group received a tailored pharmacist consultation at one of the 14 pharmacies that voluntarily participated in the study or over the phone. The consultation lasted up to 15 minutes. The pharmacist conducting the consultation had access to the women`s answers to the baseline questionnaire (Q1) in advance. This information was used to prepare a structured, individualized consultation that addressed each woman`s concerns and needs.
The control group
Women assigned to the control group received only standard prenatal care.
Data collection
SafeStart survey data
The SafeStart interventional trial included four sets of questionnaires (Q1-Q4). This study analyzed data from the first baseline questionnaire (Q1), the second follow-up questionnaire (Q2), and study pharmacist notes from the pharmacist consultation. The Q1 and Q2 were sent electronically to all women. Q1 was completed at enrollment in the first trimester, and Q2 was distributed 13 weeks after enrollment and aimed for the second trimester (Fig. 1).
SafeStart survey data – Q1
Q1 included questions about the women`s sociodemographic and lifestyle characteristics, chronic conditions, NVP severity, quality of life in the first trimester, self-reported medication use, and attitudes towards medication safety in the first trimester. Attitudes towards medication use were measured by the Beliefs about Medicines Questionnaire [24]. Q1 included also a list of health conditions (e.g., allergy, general pain, heartburn, NVP, constipation), including related medication use. Additional medication use could be reported as free text.
SafeStart survey data – Q2
The follow-up questionnaire, Q2, was distributed 13 weeks after enrollment and aimed to identify medication use in the second trimester, defined as gestational weeks 13–27. In addition to the list of medical conditions and related medication use repeated from Q1, Q2 recorded the women`s gestational week of when the Q2 was completed. This gestational week was used to identify second trimester medication use after the pharmacist intervention.
SafeStart survey data - Pharmacist notes
Pharmacist notes provided information about the consultation, such as the setting and duration, in addition to topics discussed and pregnancy-related conditions addressed during the consultation.
Prescription registry data
The SafeStart survey data were linked to the NorPD data by the women`s unique social security numbers. NorPD is a national registry covering all prescribed medications dispensed at pharmacies to individual patients in Norway living outside institutions. NorPD data includes medication name, ATC-code, defined daily dose, package size, and the dispense date to be sorted by participant. The Q1 completion date and the reported gestational week reported in Q1 were used to calculate the pregnancy start. The three months before the start of pregnancy was defined by the pregnancy start date subtracted by 90 days. The trimesters were defined as follows: First trimester: 1–90 days after the pregnancy start date, second trimester: 91–180 days after the pregnancy start date, and third trimester: 180 days after the pregnancy start date and until delivery. Three months post-partum was defined as date of delivery plus 90 days. The time point of medication exposure during the pregnancy period, which included three months before the start of pregnancy and three months post-partum, was identified by utilized dispense date as registered in the NorPD.
Data storage
All collected data were stored and analyzed at the Service for Sensitive Data at the University of Oslo (TSD) [25]. TSD is protected by a two-factor authentication, and designed for storing and post-processing sensitive data in compliance with the Norwegian “Personal Data Act”, “Health Research Act”, and regulations regarding an individual’s privacy.
Outcome measures: Medication Use
The outcome measure was medication use in the second trimester. The outcome was assessed by evaluating the differences in medication use in the second trimester among women in the intervention and the control group.
All medications were classified at the anatomical/pharmacological group by the Anatomical Therapeutic Chemical (ATC) classification system (ATC 1st level) [26]. Antiemetic medications (meclizine, promethazine, and metoclopramide) were classified at the substance level (ATC 5th level).
Statistical methods
Descriptive analyzes
We restricted the study population to women who responded to the baseline questionnaire Q1, the follow-up questionnaire Q2, and for women who received the pharmacist consultation if they were allocated to the intervention group. All analyzes were therefore performed as complete case analyzes according to the per-protocol principle.
We compared the baseline characteristics of the intervention group receiving the pharmacist consultation to the control group to evaluate whether the allocation process produced balanced groups. The chi-squared test was used to compare categorical variables, i.e., relationship status, education level, work situation, folic acid supplement, parity, pregnancy-related conditions, chronic conditions, use of self-reported medications, and filled prescriptions as registered in the NorPD. The Student’s t-test was used to compare the continuous variables, i.e., gestational week, maternal age, and Pregnancy Unique Quantification of Emesis (PUQE) score. Categorical variables are presented as counts and percentages. Continuous variables are presented as the median and range. Proportions of filled prescriptions of medications within ATC-codes with at least 20 women in the defined time periods as registered in the NorPD were calculated for the five pregnancy periods, three months before pregnancy, first-, second-, third trimesters, and three months post-partum. Filled prescriptions for antiemetic medications were considered for the first and second trimesters only.
Association analyzes
Logistic regression was performed to estimate the association between the community pharmacist consultation in early pregnancy (yes/no) on second trimester medications use, while adjusting for first trimester medication use. Separate models were computed for self-reported medication use and filled prescriptions as registered in the NorPD (medications in general and antiemetic medications). The results are presented as the crude and adjusted odds ratios (OR) with a 95% confidence interval (CI).
Sensitive analyze
We performed a pre-defined stratified analyzes according to employment status to assess effect modification by being a health care worker. We hypothesized that the intervention would have a different impact on medication use among health care workers compared to pregnant women working elsewhere as we assumed health care workers have a higher knowledge level regarding health care and medication use. All analyzes were performed with Stata/MP v.16.1.
Trial registration
The SafeStart study is registered with ClinicalTrials.gov (identifier: NCT04182750, registration date: December 2, 2019).