Determinants of healthcare providers’ confidence in their clinical skills to deliver quality obstetric and newborn care in Uganda and Zambia

Background: Poor quality obstetric and newborn care persists in sub-Saharan Africa and weak provider competence is an important contributor. To be competent, providers need to be both knowledgeable and confident in their ability to perform necessary clinical actions. Confidence or self-efficacy has not been extensively studied but may be related to individuals’ knowledge, ability to practice their skills, and other modifiable factors. In this study, we investigated how knowledge and scope of practice are associated with provider confidence in delivering obstetric and newborn health services in Uganda and Zambia. Methods: This study was a secondary analysis of data from an obstetric and newborn care program implementation evaluation. Provider knowledge, scope of practice (completion of a series of obstetric tasks in the past 3 months) and confidence in delivering obstetric and newborn care were measured post intervention in intervention and comparison districts in Uganda and Zambia. We used multiple linear regression models to investigate the extent to which exposure to a wider range of clinical tasks associated with confidence, adjusting for facility and provider characteristics. Results: Of the 574 providers included in the study, 69% were female, 24% were nurses, and 6% were doctors. The mean confidence score was 71%. Providers’ mean knowledge score was 56% and they reported performing 57% of basic obstetric tasks in the past 3 months. In the adjusted model, providers who completed more than 69% of the obstetric tasks reported a 13-percentage point (95% CI 0.08, 0.17) higher confidence than providers who performed less than 50% of the tasks. Female providers and nurses were considerably less confident than males and doctors. Provider knowledge was moderately associated with provider confidence. Conclusions: Our study showed that scope of practice (the range of clinical tasks routinely performed by providers) is an important determinant of confidence. Ensuring that providers are exposed to a variety of services is crucial to support improvement in provider confidence and competence. Policies to improve provider confidence and pre-service training should also address differences by gender and by cadres. This study uses data from the implementation evaluation of the Saving Mothers and Giving Life (SMGL) intervention [24]. SMGL has significantly improved healthcare providers’ knowledge, confidence and job satisfaction [25]. This intervention was implemented in Uganda and Zambia from January – June 2012. For each country, four districts with a high maternal mortality ratio, a low facility delivery rate, and a high healthcare provider shortage were selected for implementation (Kabarole, Kamwenge, Kibaale, and Kyenjojo in Uganda and Mansa, Lundazi, Nyimba, and Kalomo in Zambia). These districts were primarily rural, with a largely agricultural workforce. The purpose of SMGL was to increase demand for and access to facility delivery and to improve the quality of care by including activities at both the health system and community levels. Healthcare providers received training on emergency obstetric care and newborn resuscitation. The core inputs and activities of the SMGL have been previously published [26]. The evaluation methods has been described elsewhere but in brief, it used a quasi-random post-test-only comparison group design [25]. The data for the evaluation were collected from May 2013 to July 2013. The evaluation team selected four comparison districts that were similar to the intervention districts in terms of geography, health system infrastructure, health system utilization, morbidity, and mortality. The comparison districts were Masindi and Kiryandongo in Uganda and Kapiri Mposhi and Kabwe in Zambia. Health facilities with the highest delivery volumes were selected in each district. These facilities included both rural and urban and at least one referral hospital which provides comprehensive emergency obstetric and newborn care. The maternal healthcare providers (i.e. those clinicians directly involved in the provision of obstetric and newborn care such as delivery, antenatal care, and postnatal care) were invited to complete three questionnaires – a clinical confidence and scope of practice questionnaire, a job satisfaction questionnaire and an obstetric knowledge test. Eligible healthcare providers included all clinicians (i.e., doctors, nurses, midwives, nurse assistants, and clinical officers) who worked in maternity wards or provided obstetric and newborn care. All questionnaires with the exception of the knowledge test were conducted face-to-face; the test was self-administered. The knowledge test was adapted from one previously developed and validated by Johns Hopkins Program for International Education in Gynecology and Obstetrics (JHPIEGO) [27] a maternal health program based at Johns Hopkins University. This test included 60 multiple-choice questions on general obstetric knowledge for the management of early pregnancy, labor and delivery, and postpartum care. The knowledge test was developped based on World Health Organizations’ international guideline Managing Complications in Pregnancy and Childbirth: A Guide for Midwives and Doctors [28]. This guideline has been used throughout Africa, Asia, and the Americas [25]. Prior to start of data collection, the three instruments were pilot tested in non-study districts in Uganda and Zambia, and revised accordingly [25]. Providers’ demographics and facility characteristics were also included in the job satisfaction questionnaire. We created a confidence score based on providers’ own assessment of their ability to perform 27 obstetric and newborn care tasks including for example, administering oxytocin, repairing perineal tears, and performing Kangaroo care (See Additional file 1 – Appendix E). The primary study [25] used a confidence score that included 26 tasks. For the present study, we decided to add two others “Newborn Apgar assessment” and “Kangaroo care” since these tasks are important component of immediate newborn care. We also removed “Administering anesthesia for C-sections” due to having only one anesthetist in our study sample (N = 1). Providers rated their confidence in being able to perform each task as very confident, not very confident, I cannot perform this skill, and does not apply. Providers who rated themselves as ‘very confident’ in performing a task were given one point per task and zero for ‘not very confident’ and ‘I cannot perform this skill’. Provider confidence was only assessed for tasks that each cadre believed they were expected to perform. ‘Does not apply’ responses were therefore treated as missing. For example, only general doctors, specialists, or medical licentiates [29] are expected to perform C-sections. The total score was converted into an average out of 100. Scope of practice was measured by asking providers whether they had performed these same 27 tasks in the past 3 months. The scope of practice score was based on 27 tasks for doctors, obstetrics/gynecology specialist, and medical licentiates and 26 tasks for other cadres (performing C-sections was excluded for the other cadres). The total score was converted into a percentage. The average provider knowledge score was calculated out of 100. We categorized the score into tertiles, after observing non-linearity in bivariate association between knowledge and confidence. The resulting categorical variable included providers with low (28–50%), average (51–58%), and high (60–93%) knowledge scores. As with knowledge score, we categorized this into tertiles after observing non-linearity in the association between scope of practice and confidence. The resulting categorical variable included providers with low (0–48%), average (50–65%), and high (69–100%) levels of clinical practice in the past 3 months. Other variables of interest included age, days of training in the past year, gender, qualification, the facility type, public or private ownership, whether the provider was part of the intervention or control group, and country. Age was treated as a continuous variable. A quadratic term for age was also included to account for non-linearity. Amount of training received in the past year was measured as the total number of days during which providers reported receiving on-site trainings. There were seven cadres of providers in the study districts with different levels of training: 1) nurse assistant, 2) enrolled nurse, 3) enrolled midwife, 4) registered nurse, 5) registered midwife, 6) clinical officer and 7) general doctor, doctor specialists, and the medical licentiate. Nurse assistants are trained for about 6 months and exist only in Uganda. Enrolled nurses and enrolled midwives are trained for 2–3 years. Enrolled nurses are similar to licensed practical nurses in the United States. Registered nurses and registered midwives receive medical training for 3 to 4.5 years. Clinical officers receive 3 years of training. Doctors are typically trained for 5 to 7 years. Medical licentiates are clinical officers who received additional training so that they can perform the tasks that a doctor would typically perform [29]. Given the similar years of training, medical licentiates were included in the group of doctors. Facility type was based on the availability of services at the facility and categorized into two groups. The first one included health centers that provide basic emergency obstetric and neonatal care (BEmONC). Seven signal functions are provided in BEmONC facilities: 1) Administration of parenteral antibiotics, 2) Administration of uterotonic drugs for active management of the third stage of labor and prevention of postpartum hemorrhage, 3) Use of parenteral anticonvulsants for the management of preeclampsia/eclampsia, 4) Manual removal of placenta, 5) Removal of retained products, 6) Assistance of vaginal delivery, and 7) Basic neonatal resuscitation [30]. The second category included hospitals that provide comprehensive obstetric and neonatal care (CEmONC) which perform the 7 basic functions of BEmONC and two additional services: cesarean delivery and blood transfusion [30]. Private ownership included both for-profit and not-for-profit facilities. We used a multiple linear regression model to investigate the factors associated with confidence, adjusting for covariates at the facility and provider levels. The unit of analysis was the provider and the regression model adjusted standard errors for clustering at the facility level. Three sensitivity analyses were performed. First, we added an interaction term between knowledge and scope of practice to observe whether the association between knowledge and confidence differed across levels of practice. Second, we regressed confidence on continuous measure of knowledge and scope of practice rather than tertiles. Third, we repeated the analysis in the intervention and the control facilities separately. All analyses were conducted in September 2019 using Stata SE version 16.0. P-values lower than 0.05 were considered statistically significant. Since this study was a secondary analysis of de-identified data, it was not deemed to be human subjects research (NHSR) under the Harvard T.H. Chan School of Public Health Institutional Review Board (IRB) policy [31]. Thus, NHSR is exempt to acquire a formal IRB determination. The original study was approved by IRBs at Columbia University in the United States, Makerere University and the National Council for Science Technology in Uganda, and Excellence in Research Ethics and Science Converge Research Ethics Committee and Ministry of Health in Zambia [25]. Consent was obtained from the health care provider interviewed and the job satisfaction and knowledge surveys were completed in private rooms to ensure privacy. Data used for the study were stored in a secured folder with limited access.