Introduction

Unintended pregnancies are a global burden in healthcare. The rate of unintended pregnancy, although in decline, is still high in developing countries [1, 2]. Emergency Oral Contraception (EOC) is one of the methods proposed for the prevention of unintended pregnancies [3, 4]. Unfortunately, women with unintended pregnancies usually do not use or are not aware of EOCs. Studies showed that inadequate knowledge of emergency contraceptives in users was associated with unplanned pregnancy [5,6,7]. Cultural beliefs, fear of side effects, lack of knowledge, and religion are also the reasons for refusing to use and provide EOCs [8].

The Knowledge, Attitudes, and Practices (KAP) theory states that knowledge and attitude are important factors in changing behaviors or practices [9]. Providing correct information on EOCs should lead to improving attitudes, thus improving behaviors that are related to EOC acceptance. However, a weak correlation between KAP has been criticized [10], so having correct knowledge of EOCs might not lead to positive attitudes or EOC use. In EOC users, this might not be the case, since literature shows that providing EOC knowledge to EOC users improves EOC use [11]. However, such information in healthcare professionals does not exist. Therefore, KAP in healthcare professionals is important to understand the barriers to EOCs from the provider’s end. Pharmacists increasingly involve with EOCs, since levonorgestrel EOCs are provided to patients by pharmacists without the need for prescriptions in several countries, e.g., Australia, Canada, China, Thailand, USA, and the UK [8]. Consequently, investigating KAP related to EOC in pharmacists is important for EOC promotion.

In Thailand, levonorgestrel EOCs are also available in pharmacies, although users can access the EOCs from hospitals or clinics. We assume that most of the EOCs are provided in pharmacies and users request EOCs directly from community pharmacists, albeit this has never been confirmed. Information on KAP related to EOCs in Thai pharmacists is not available [8], although Thai people may stereotypically perceive that KAP in Thai pharmacists is poor. For example, in the short film, Patcha is Sexy, a pharmacist refused to provide EOC for a teenager and a pharmacist replaces EOCs with a look-alike drug in the series, You Are My Makeup Artist. In this study, we conducted a cross-sectional online survey of Thai pharmacists to explore their baseline KAP. The correlation of KAP related to EOCs and factors that explained the KAP were also investigated.

Methods

This research consisted of three major parts. The first one was to collect the necessary information for the KAP survey construction. We conducted in-depth interviews, Google Trend searches, and Pantip.com searches to collect the data in this phase. The second part was to design and evaluate KAP survey questions. To construct the KAP survey, we compared data from the first part with data from a previously published systematic review and meta-analysis [8]. These questions were then evaluated by experts. In the last phase, KAP cross-sectional survey was distributed to collect data on KAP related to EOCs and the correlation among KAP components.

Collecting data for survey construction: in-depth interview

The protocol for the research was approved by Burapha University Institutional Review Board (IRB1-003/2565). This in-depth interview complied with the Consolidated Criteria for Reporting Qualitative Research. The first author is a registered pharmacist with an assistant professorship and had published a few qualitative studies in peer review journals. He trained 3 year five female pharmacy students to be Research Assistants (RAs). No assumptions were assumed, and the RAs were not personally interested in the topic. Phenomenology was used in this study. The first author and his RAs designed semi-structured interview questions for pharmacists. The questions were also reviewed by two independent registered pharmacists. The interview questions were pilot-tested on five pharmacy students and modified accordingly (Additional file 1: Data 1).

The interview occurred from October 2021 to March 2022. The participants were selected by convenience sampling and approached face-to-face. The target number of enrollments was approximated at 25 [12] or when the theme was saturated. Currently registered pharmacists with active pharmacist licenses were included. The interview was scheduled later online via GoogleMeet (Google LLC, USA) because of the coronavirus disease pandemic. Participants were free to select their interview place. No relationship between researchers and interviewees was established before the interview. However, participants knew that the interview was required for the graduation of the RAs. The RAs conducted and audio-recorded all interviews. Field notes were taken during the interview.

The RAs created transcripts from the interviews. The first author and the RAs read and discussed the transcript together. Comments on the transcript were generated in Microsoft Word 365 (Microsoft Corporation, USA). The comments were categorized and collected as themes by consensus. Themes were used to further improve the interview questions until no new themes emerged. The transcript was not returned to the interviewees, so no feedback was obtained. The themes and illustrative quotes were translated into English by the first author.

Collecting data for survey construction: Google Trend searches

Words related to EOCs were searched in Google Trend (Google LLC, USA) from 2004 to February 2022. Search terms in Thai including emergency oral contraceptive, and morning-after pill were searched by limiting the location in Thailand, while search terms in English (emergency contraception (pill), morning-after pill, levonorgestrel, Postinor, and ulipristal) were searched worldwide. Related queries retrieved from each search were compiled in Microsoft Excel 365 (Microsoft Corporation, USA). The list of queries was generated after redundancy was removed.

Collecting data for survey construction: Pantip.com searches

https://pantip.com/ is the largest online Thai community since 1997. The website, although it is not a search engine, has a function that allows searching for related topics on the webpage and its forum categories that are discussed. However, the maximum number of retrieved forums is 20,000 per search. We used a Thai search term for morning-after pills to search from inception to April 1st, 2022. Questions and concerns related to EOCs were manually collected, transferred to Microsoft Excel 365, and categorized.

Survey question design and evaluation

The protocol for the research was approved by Burapha University Institutional Review Board (IRB1-085/2565). Data from the in-depth interview, Google Trend searches, Pantip.com searches, and a previously published systematic review and meta-analysis [8] were used for the design of the KAP survey. We aimed to collect data on the demographics, and EOC-related KAP of pharmacist participants. Knowledge questions were closed-ended with multiple-answer choices. Self-rating KAP; self-rating general attitudes (e.g., religiousness, openness to sex discussion); attitudes; and practice questions were answered as five-point Likert scales. The questions were reviewed and modified by all authors. To assess content validity, the survey questions were rated by five, a recommended number from the literature [9], non-author pharmacist experts using a five-point Likert scale, where 5 = highly relevant, 4 = relevant, 3 = neutral, 2 = mostly not relevant, and 1 = totally not relevant. Then, the index of Item Objective Congruence (IOC) was calculated for each question. In addition, the expert panels also checked and commented on the appropriateness of survey questions and their response options. After the modification according to the expert comments, the third author transformed the survey into Google Forms (Google LLC, USA.), which could be accessed only by the first and third authors. Google Forms was set to not include multiple submissions from the same individual. The survey was sent to 10 pharmacy students for their opinion on the comprehensibility of the questions and the feasibility of the online survey.

Data collection via KAP cross-sectional survey

The number of registered pharmacists reported by the Thai Pharmacy Council on December 1st, 2021, was 45,707. We use Open Epi online version 3.01 to calculate the sample size using its default setting and alpha of 0.05. With a 10%-excess number of participants, we aimed to collect data from 420 pharmacists. We included all pharmacists of any age with an active pharmacist license who were willing to answer the survey. The advertising material was posted in online communities of Thai pharmacists, e.g., a Facebook page Pharmacafe with approximately 26,800 members, and groups of pharmacists in the Line application, to ensure that we reached the target samples. We recommended participants answer the survey questions online on their computers to reduce fatigue. This survey was reported according to A Consensus-Based Checklist for Reporting of Survey Studies.

Data and statistical analysis

Data were exported and cleaned in an Excel worksheet. Missing data and non-response errors were assumed to be missing completely at random and were handled by exclusion. Imputation and adjustment for the non-representativeness of the sample were not conducted. We defined total scores as a sum of correct answers from selected questions regarding knowledge, attitude, or practice, while self-rating scores were 5-scale Likert scores from where the participants rated themselves on their knowledge, attitude, or practice. Attitude and practice questions that were used for calculating the total attitude and total practice scores were discussed and reached by consensus among authors.

Descriptive statistics were performed using Microsoft Excel 365. Inferential statistics were conducted using SPSS (IBM SPSS Statistics 28.0.0.0, IBM, USA). We planned a priori to see the correlation among KAP by calculating Spearman’s rank correlation coefficient. The coefficient of less than 0.4 and 0.7 were classified as weak and medium correlations, respectively [13]. Associations between EOC dispensing practice and 17 independent variables including demographics (age, degree, marital status, opinion on unintentional pregnancy, practicing duration, type of pharmacy practice, religions, self-rating judgment, self-rating religiousness, self-rating openness to sex discussion, and sex); a total score and self-rating KAP, were analyzed by the linear mixed model. The number of Subjects per Variable (SPV) in this study was approximately 24 (420/17) which was higher than 10, the recommended lowest SPV [14]. No adjustment for the non-representativeness of the sample and sensitivity analysis was conducted.

Results

Overall, 64, 26, 75, and 115 questions emerged from the in-depth interview, Google Trend searches, Pantip.com search, and the previously published systematic review and meta-analysis [8], respectively. After eliminating repeated questions, we obtained 57 knowledge-related questions and 103 attitude-related questions. The questions were evaluated by the experts with an IOC of > 0.7. The survey was sent to registered Thai pharmacists. Four-hundred and twenty-seven responses were obtained and used for descriptive and linear-regression statistical analysis.

Collecting data for survey construction: in-depth interview

We included 17 pharmacists in the in-depth interview. Although we did not reach the planned sample size because of the heavy lockdown from the coronavirus disease, we reached theme saturation. All the pharmacists that we reached out to agree to participate in the interview. All interviews were conducted online by all three RAs without the presence of other people. The median interview duration was 15 min (interquartile range 5–15 min). No repeated interview was conducted. The median age of interviewees was 35 years, and the median practicing duration was 3 years. Most of the interviewees were female, Bachelor’s degree graduates, community pharmacists, and Buddhists (Table 1).

Table 1 Demographics of pharmacists who participated in the qualitative part of this study
Full size table

All themes are illustrated in Table 2. Briefly, Thai pharmacists agreed that sex and unwanted pregnancy is still a very sensitive topic among patients and pharmacists. In addition, pharmacists agreed that sex and EOC education should be provided at a young age and families should be involved in providing the education. Some pharmacists perceived that using EOCs was an act of responsibility, while some still had negative attitudes toward EOC use. Some pharmacists expressed concern had their children used EOCs. Interestingly, inaccurate information on contraception and EOCs was found during the interview.

Table 2 Theme emerged during the semi-structured in-depth interview
Full size table

Collecting data for survey construction: Google Trend searches

We found 665 queries about levonorgestrel and ulipristal acetate EOCs in EOC users. After removing redundant queries, 221 EOC-related queries were categorized into 34 topics and used in 26 survey questions (Tables 3, 4, 5).

Table 3 Survey questions for EOC knowledge and their sources, answers, references for the answers, and scores
Full size table
Table 4 Survey questions for EOC attitude and their sources, the frequency of the attitudes, and possible scores
Full size table
Table 5 Survey questions for practice during EOC dispensing and their sources, the frequency of the practice, and possible scores
Full size table

Collecting data for survey construction: Pantip.com searches

We accessed the latest 2000 forums from November 8th, 2020, to April 1st, 2022, and found 82 topics which were then used in 75 survey questions (Tables 3, 4, 5). In addition, text analysis allowed us to extract themes related to EOC knowledge and attitude in Thai internet users. First, sex literacy was low and sex education was inadequate. As a result, the internet was a main source of sexual health information. Second, incorrect information and wrong understanding, e.g., the limited number of EOCs per lifetime, permanent harm from EOCs, coitus interruptus and fertility awareness method as effective contraception, and vaginal bleeding as an indicator for pregnancy, were prevalent.

Cross-sectional KAP survey

The final cross-sectional survey included 186 questions and was classified into 5 sections. There were 10, 57, 103, and 16 questions for demographics, knowledge, attitudes, and practice, respectively (3, 4 and 5). All of them had an IOC > 0.7. From October 2022 to January 2023, we received 427 responses, but 421 responses were included after data cleaning. The demographics of the participating pharmacists are summarized in Table 6. Briefly, most of the participants were adults; female; graduated with B.Pharm. or Pharm.D.; community or hospital pharmacists; in private section; Buddhists; and single. They were also averagely religious, open to sex discussion, and low or moderately judgmental. More than 90% of them agree that unintentional pregnancy is a social problem (Table 7).

Table 6 Demographic data of survey participants
Full size table
Table 7 Descriptive statistics for knowledge, attitude, and practice regarding levonorgestrel emergency oral contraceptive
Full size table

Pharmacists’ knowledge of EOCs

All participants were aware of EOCs and 48.69% of them reported good knowledge about EOCs (median knowledge scores = 51.02%; Table 7). We found that, first, the total knowledge score was weakly correlated with the self-rating knowledge score and was not a predictor of the self-rating knowledge score. Second, both total and self-rating knowledge scores were weakly correlated with attitude and practice scores. However, the self-rating knowledge score was moderately associated with the self-rating practice score (ρ = 0.525, p < 0.001, Table 8). In addition, the total knowledge score and self-rating knowledge score could explain variances in the total attitude score and self-rating practice score but not the self-rating attitude score and total practice score. Third, graduating with a clinical pharmacy degree, and practicing community or hospital pharmacy was associated with an increased total knowledge score while practicing community pharmacy, not being single, and being open to sex discussion was associated with an increased self-rating knowledge score (Table 8).

Table 8 Generalized mixed model for the knowledge, attitude, and practice
Full size table

In this study, less than 50% of the pharmacists knew the correct answer for the tradename, maximum doses for continuous use, the window period for repeating doses after vomiting, efficacy reduction after frequent use, the effective period after a complete dose, teratogenicity, relationship with abortifacient action, and drug interaction. In addition, less than 25% knew the correct answer for the mechanism of action, maximum doses per month, effectiveness after ovulation, the window period for side effects after a complete dose, serious side effects that required EOC cessation, and details in menstrual disturbances (Table 3). Moreover, the knowledge of EOC contraindications should be questioned since 62.94% and 64.13% of the participants knew that breast cancer and severe hepatic impairment, respectively, was a contraindication for levonorgestrel, while 66.50%, 46.32%, and 25.65% of the participants misunderstood that thromboembolism, ectopic pregnancy, and hypertension, respectively, was the contraindication (Table 3).

Pharmacists’ attitude toward EOCs

Most of the participants (48.22%) reported a somewhat positive attitude toward EOCs (median attitude scores = 21.81%; Table 7). We found that, first, the total attitude score and self-rating attitude were weakly correlated (ρ = 0.307, p < 0.001) (Table 8). However, the total attitude score significantly explained the self-rating attitude score (p < 0.001), and vice versa (p < 0.001). Second, the total attitude score did not explain variances between the total practice score (p = 0.909) and self-rating practice score (p = 0.164), but the self-rating attitude score explained the total practice score (p = 0.013) and the self-rating practice score (p < 0.001; Table 8). Finally, practicing community or hospital pharmacy, being religious, and not being single were associated with a lower total attitude score, while being non-Buddhist was associated with a lower self-rating attitude score (Table 8).

Attitudes expressed by the participants were summarized (Table 4). Pharmacists agreed that (1) EOCs should not be used with the lactational amenorrhea method and combined oral contraceptives because of safety concerns; (2) EOCs should be used when ejaculation occurred; (3) EOCs were a fast-acting, effective, and dangerous contraception alternative; and (4) EOC knowledge should be promoted in anyone at puberty ages, confirming the themes found during the in-depth interview. Although pharmacists did not agree with EOC use restrictions based on age, they felt that selling EOCs in schools was not appropriate, while selling EOCs in universities was acceptable.

Data from both the in-depth interview and survey showed that, to pharmacists, using EOCs showed responsibility. They perceived that EOC users were shy and had negative feelings toward EOC use, which could be a barrier to EOC use. Most pharmacists did not feel that they were the barrier to EOC use or that EOCs should be prescription drugs. Pharmacists felt that dispensing EOCs was their responsibility, and selling EOCs should be under their supervision. Most of them did not approve of EOC refusal because of personal reasons. However, EOC refusal based on medical reasons was acceptable. This finding agreed with the qualitative study.

From pharmacists' viewpoints, package inserts, media, partners, parents, friends, and healthcare professionals affected the decision-making of EOC users. Pharmacists felt that responsible male partners would have active roles in obtaining knowledge and making the decision on EOC use. Eighty-five percent (356/417) of the pharmacists reported that religion did not affect their decision on EOC dispensing. Less than 3% of Thai pharmacists perceived using EOCs as immoral or sinful. However, some of them felt that religion did affect whether patients used EOCs or not. Although pharmacists' attitudes toward EOC use and promiscuity or other sex behaviors were neutral, they perceived that EOC use increased the risk of sexually transmitted diseases.

Pharmacists’ practice on EOCs

Most of the participants (56.77%) reported that they had good pharmacy practice during EOC dispensing (median practice scores = 60%; Table 7). We found that, first, the correlation between the total practice score and self-rating practice score was significant but very weak (ρ = 0.148, p = 0.003). The total practice score did not explain the variances of the self-rating practice score and vice versa. Second, practicing community pharmacy was associated with decreased total practice score, while being more open to sex discussion and not seeing unwanted pregnancy as problems were associated with an increased total practice score. In addition, being younger, being non-Buddhist, defining oneself as religious, and viewing unwanted pregnancy as problems were associated with an increased self-rating practice score (Table 8).

More than 80% of pharmacists reported that they dispensed EOCs, because their patients requested the EOCs. Approximately 40% of them reported that they never refuse to dispense EOCs. The major reason for EOC refusal is contraindications. In addition, 60% of them reported they would not have dispensed EOCs for the patients had the patients already used 3 mg levonorgestrel EOCs within that month. Pharmacists had very diverse practices of asking for the reason for EOC requests, ranging from never asking to always asking. Half of the participants would recommend switching EOCs to combine oral contraceptives and recommend concomitant condom use. Pharmacists frequently emphasized to their patients to take EOCs on time or suggested their patients set an alarm. Sending reminder messages was rarely practiced (Table 5).

Discussion

In Thailand, levonorgestrel EOCs can be sold by pharmacists without prescriptions since 1997 (the Order of Ministry of Public Health 1037/2543). Contraception is also listed as one of the core competencies for Thai pharmacists since 2002 [15]. The long experiences with EOCs and compulsory education on contraception may lead to the 100% EOC awareness found in this study, which was higher than what was reported in a meta-analysis [8]. However, this did not change the fact that EOCs are still misunderstood and misperceived by pharmacists, as supported by findings from the interview and the survey. The knowledge scores were not as high as reported in the previous study [8], as we expected, since we asked pharmacists highly detailed oriented real-world questions from EOC users. In addition, most of the participants reported a neutral or positive attitude toward EOCs. In this study, refusing to dispense EOCs because of nonmedical reasons was perceived as “unacceptable”. However, pharmacists imprecisely identified thromboembolism, ectopic pregnancy, and hypertension as levonorgestrel contraindications. Misunderstanding the contraindications of EOC can potentially pose a barrier to EOC access.

We found that the self-rating KAP scores were higher than the total KAP scores. In addition, using total scores or self-rating scores led to different conclusions on the relationship between KAP. For example, the total practice score could not be predicted by either the total knowledge score or the total attitude score, but the self-rating practice score could be predicted by self-rating knowledge and self-rating attitude. Several other KAP surveys used a summation of correct answers among knowledge, attitude, and practices [16,17,18], and a guideline to conduct a KAP survey also recommends using the total score [9]. We used both approaches and found that how pharmacists perceive their knowledge and attitude toward EOCs affected how they perceive their practice, but the knowledge and attitude did not affect their practice.

Several variables affect the total scores of KAP. First, pharmacists who practiced in pharmacies and hospitals had a higher total knowledge score than those who practiced in other settings (e.g., manufacturing plants or regulatory offices). A degree in clinical pharmacy is also associated with higher knowledge scores, agreeing to a study in Nepal [19]. Second, even though religion may have affected self-rating attitude or practice scores, it did not affect total KAP scores. This agrees with a study on Christian pharmacy students which also found that religion did not affect the attitude toward EOCs [20]. Third, the year of practice, which was found to be associated with dispensing practice [21], did not explain the practice in our study. This might be because all registered pharmacists in Thailand must participate in annual Continuing Pharmacy Education (CPE).

The major limitation of this study was that the survey was very long. Other limitations due to the cross-sectional nature of this survey were that, first, this survey might not accurately reflect temporal variation in pharmacists' KAP. Second, recall bias may occur with some questions that require recollection of the data retrospectively. Third, using online surveys can lead to the selection of younger pharmacists who are more compatible with using the internet. Despite these, our study has several implications. We have several positive feedbacks suggesting that distributing the survey answers (Table 3) can help correct EOC misunderstandings and can be helpful in pharmacy education and training. For other applications, we provide preliminary evidence that the relationship between KAP yielded from the total scores and self-rating scores can disagree. In addition, even when the EOC can be accessed without prescription, the awareness reaches 100%, and the self-reported attitude toward EOCs among pharmacists is positive, myths that prevent EOC access, especially those regarding contraindications, can be prevalent. This highlights the importance of providing updated knowledge relevant to EOC among Thai pharmacists through a variety of channels including CPE. For further application of this study, since education and economics affected EOC awareness [8], this study should represent Buddhist countries with middle levels of income and education, where EOCs can be accessed without prescriptions.

Conclusions

Most of the access to EOCs in Thailand is derived from patient requests to pharmacists. Therefore, Thai pharmacists play important roles in the care and promotion of EOC use. We found discrepancies between self-rating KAP scores and total KAP scores. Pharmacists rated their KAP scores higher than their actual KAP scores. The correlation among KAP components was weak. Self-rating scores showed that knowledge and attitudes were associated with the practice, but the total scores showed that the knowledge of or attitude toward EOCs was not associated with EOC practice. EOC knowledge and attitudes should be promoted to prevent refusal of EOC dispensing but this may not improve EOC practice.