Clinicians’ knowledge and practices regarding family planning and intrauterine devices in China, Kazakhstan, Laos and Mexico

The questionnaire administered in this study was developed based on nine existing questionnaires [34‐42]. Development and testing of the questionnaire showed it to have high internal consistency and good face and content validity. Details on its development, pilot-testing, assessment for validity and reliability, and translation have been documented elsewhere [28]. The survey tool incorporated questions addressing participants’ individual characteristics, working context, training, networking activities, and access to, trust in and use of research evidence. Additionally, five true/false questions were used to test respondents’ knowledge of IUDs for family planning and five other questions assessed relevant clinical practices [43]. The questionnaire was translated into Mandarin, Russian, Lao and Spanish by WHO’s translation service and local country teams [30].

Country teams administered the questionnaire in China, Kazakhstan, Laos and Mexico between October 2004 and December 2005. The teams were each led by one co-author (GS in China, BST in Kazakhstan, BB in Laos, and FBP in Mexico) and supported by the Research to Policy & Practice Study Team (all members listed in Acknowledgements section). The four selected LMICs differ in population size, per capita income, health expenditures, life expectancy, contraceptive prevalence, and Internet access (Table 1) [44‐49]. Each team aimed to gather complete responses from a representative sample of at least 100 clinicians in each country (the vast majority of participants were physicians). The questionnaire was administered using a drop-off and pick-up approach (except in Mexico, where a mix of drop-off and pick-up and in-person administration was utilized). Various approaches to increase response rate were employed, including personalized letters and providing a free set of WHO publications as an incentive. More detailed information about the data collection process is described elsewhere [28, 29].

Systematic reviews are credited as the ideal method of synthesizing global research evidence and offering summary information to aid decision-makers. Furthermore, they are extensively available and internationally authoritative on clinical interventions like IUDs [50‐52]. As a result, systematic reviews have been extensively used in this analysis, with the Cochrane Library representing the most comprehensive identified source. Efforts to distinguish between summaries and full reports, as well as between scientific journals from a high-income country and a participant’s own country, were guided by past studies [29, 36, 53].

In China, the team used a stratified random sampling process (with a sampling frame constructed from an existing list of family planning centres) to sample 120 clinicians who provided care to women seeking contraception. Stratification was by geographic location (Sichuan province in the southwest, and Liaoning province in the northeast) and by type of facility (Maternal and Child Health clinics, Departments of Obstetrics and Gynecology in hospitals, and Family Planning service institutions).

In Laos, a sampling frame was constructed from existing lists of clinicians retrieved from the Ministry of Health’s Department of Human Resources for Health and four provincial health departments, and was stratified by facility type (central, provincial and district hospital). 106 clinicians providing contraceptive care in the capital city of Vientiane and the provinces of Vientiane, Borikhamsay and Savannakheth were sampled.

In Mexico, the team used a sampling frame constructed from five sources (physicians working for the Mexican Foundation for Family Planning, physicians working for the Instituto de Seguridad y Servicios Sociales de los Trabajadores del Estado, physicians working for the Secretaria de Salud [Ministry of Health] in five facilities of Mexico City, physicians involved in family planning attending a training course at the Desarrollo Integral de la Familia and at the Instituto Mexicano del Seguro Social, and private physicians involved in family planning attending a training course). The sampling process was used to sample 122 clinicians in Mexico City and the states of México, Nuevo León and Jalisco.

Using relevant items from the data collected, basic descriptive statistics were calculated and simple ordinal logistic regressions were conducted. The regression models were used to analyze the factors associated with clinicians’ knowledge and practices on IUD use and family planning. Composite knowledge and practice scores were constructed for each respondent and used as the dependent variables in the regression analyses. Knowledge scores were based on the proportion of the five true/false knowledge-testing questions that were correctly answered. Each question was assigned equal weight and no penalty was given for an incorrect answer. Similarly, practice scores were based on the frequency in which a clinician reported performing the five practices on a five-point scale (i.e., 1 “never”, 2 “rarely”, 3 “sometimes”, 4 “often”, and 5 “very often”). The scale was inverted for one described practice that is contrary to recommended practice. Each question held equal weight to ensure practice scores were integers that ranged from five to 25.

Knowledge and practice scores were converted to tertiles for inclusion as ordinal variables. For both knowledge and practice models, independent variables included clinicians’: 1) utilization of specific sources of research evidence; 2) views and behaviours related to the improvement of clinical practice; and 3) individual and practice characteristics. In order to substitute for missing values, multiple imputation using multivariate normal regressions was employed (with 100 imputations). Observations with missing dependent variables were omitted from the analysis. Stata/MP 11.2 for Mac was used to perform all statistical analyses [54].

The findings from this study suggest that significant gaps exist in clinicians’ knowledge and practices relating to family planning and IUD provision. From the population surveyed, only 12 of 434 (2.8 %) clinicians were able to correctly answer all five knowledge-testing questions. Similarly, only four of 436 (0.9 %) clinicians reported to “often” or “very often” perform the four recommended practices and never perform the one practice that was not recommended. These findings of less-than-optimal clinical knowledge and practices match those found in previous studies [19‐27]. These gaps in IUD provision and family planning, however, may be a symptom of broader gaps in clinicians’ knowledge and practices for various health issues across low-, middle-, and high-income countries [55‐71].

Through analyzing these knowledge and practice gaps, this study highlighted numerous factors associated with clinicians’ knowledge and/or practices related to family planning and IUD provision. Previous research findings have indicated that clinical training in IUD insertion can result in higher numbers of IUD insertion procedures [17] and fewer misconceptions about IUDs [21]. Accordingly, IUD training for clinicians will likely help to facilitate overall better practices in contraceptive and IUD provision in LMICs. Since gaps in knowledge and practices have varying degrees of consequences – with each gap resulting in more or less serious health and health system consequences – interventions targeting these gaps should be prioritized in whatever way is likely to produce the greatest health and social impact. Documented examples of knowledge and practice gaps and their health systems implications include the significant loss of life resulting from sub-optimal use of oral rehydration therapy for diarrhea and insecticide-treated bed nets for malaria prevention [3].

The ordinal logistic models presented in this study indicate that the knowledge and practices of clinicians may be disproportionately affected by certain factors more than others. Though it is possible that knowledge and practices are affected by different factors, the two regression analyses may also have been affected by social desirability biases. These biases would have likely affected the self-reporting of practices more so than in the answers to questions testing clinicians’ knowledge. Other explanations for the differing results could be the presence of confounders skewing the analysis, the models might have lacked necessary statistical power, and/or the insufficiency of the composite scores in representing clinicians’ actual “knowledge” and “practices”.

There are five notable strengths of this study. The first is the diverse collection of data from four distinct LMICs which differ in characteristics such as life expectancy and contraception coverage rates. Second, the data collection from all four countries yielded very high response rates. Third, the data collected and examined is about the use of IUDs, a family planning method for contraception recommended by WHO and noted as essential for achieving global health goals [4, 5]. Fourth, the questionnaire was adapted from existing tools and assessed for reliability and validity [28]. Fifth, the knowledge and practice scores used in the analysis were calculated from a range of testing questions for which respondents were not given the correct answer. This metric is more objective than relying on the participating clinicians to self-evaluate whether they had “high” or “low” knowledge and/or practices – a method used in previous studies in this field.

Despite these strengths, the study also has four notable limitations. The first relates to the professional translation of study instruments into Mandarin, Russian, Lao and Spanish. It is possible that linguistic and cultural differences in the four countries could have affected participants’ understandings of particular translated survey questions. Second, the knowledge and practice scores were calculated from responses to only ten questions. Third, the questionnaire requested self-reported data to assess providers’ practices – a metric which is subject to social desirability bias. Self-reporting of practices may overestimate actual behaviour, and correspondingly result in an over-reporting of recommended practices [72]. Finally, the study faced resource constraints, causing a long delay in data analysis and manuscript writing and limiting the survey’s representative samples of the types of clinicians in each LMIC. For instance, the sample surveyed in Kazakhstan consisted of gynaecologists working in family planning, whereas respondents in Laos consisted of any clinician providing contraceptive care. As a result, the data obtained cannot be compared across the four countries. These variations also likely affected intra-country responses and, consequently, the data and results presented are not generalizable within country. The use of uniform study tools among uniform types of clinicians across each country would likely yield different results than those observed. To strengthen generalizability, studies resembling the current analysis should be conducted using representative samples of the same types of clinicians across countries.

Clinicians serve as primary disseminators of health information and consequently hold considerable influence over patients’ health-related behaviours [1, 11‐15]. In fact, recent literature surveying women’s attitudes toward contraception has revealed that physicians may be the source of greatest influence in the decision of which type of contraception is used [73]. With an acknowledgement of the significant impact that clinicians have on individuals and public health overall, efforts to improve clinicians’ knowledge and practices facilitate an opportunity to enhance family planning and maternal health globally. With the present study’s identification of gaps in clinicians’ knowledge and practices in this field, corresponding interventions to target such gaps are needed to help reduce the number of maternal deaths and contribute to the improvement of women’s reproductive health worldwide. With the implementation of programs focused on clinicians, it is possible that comparatively small investments could produce significant returns among larger populations. These programs should be informed by the large corpus of research on clinician training and behavioural change, including: educational meetings [74]; reminders [75]; audit and feedback [76]; outreach visits [77]; distribution of educational materials [78]; local opinion leaders [79]; and the freely available Health Systems Evidence database which contains these types of reviews [80]. Although, of course, even with the best training, clinicians would face many system-level barriers to optimal family planning services and IUD provision – including lack of availability, poor quality and high costs – which means that interventions targeting clinicians must be complemented by other interventions to strengthen the health systems in which they work. That is, policies and programs towards pressing health concerns, such as family planning and IUD provision, should be informed by evidence which considers the multifactorial interactions, as well as limitations, of the broader health system in which they are to be implemented [81].

Future studies seeking to contribute to an understanding of clinicians’ knowledge and practices on family planning and IUD provision should seek to include representative samples of similar types of clinicians to enhance the generalizability of their findings across countries studied. In particular, funding to conduct research across larger, nationally representative samples of clinicians would allow for illustrative cross-sectional analyses. While this investigation has highlighted various factors that are associated with better knowledge and practices among clinicians, there remains the need for more evaluation and implementation of effective interventions to close existing know-do gaps.