Characterizing Epilepsy Treatment and Patient Outcomes in Uganda: A Prospective Hospital-Based Cohort Study

Background: Epilepsy is one of the most common neurological disorders globally, and an overwhelmingly high number of people who suffer from epilepsy reside in low- and middle-income countries (LMICs), where access to care is a significant challenge. Stigma associated with epilepsy and poor knowledge of the disease further hamper access to appropriate care. According to the World Health Organization (WHO) treatment guidelines, epilepsy can be treated cost-effectively in low-resource settings. However, in many LMIC settings, capacity for epilepsy treatment is inadequate. In Uganda specifically, there is a significant treatment gap in epilepsy care, and the magnitude of and reasons for this gap are not well characterized. In addition to suffering caused by disease in people with epilepsy (PWE), there are severe social ramifications of disease due to stigma. Therefore, access to adequate treatment is imperative for PWE in Uganda.

Purpose: The general objective of the overall project is to characterize epilepsy treatment and patient outcomes in public hospital-based care in Uganda, specifically in Mulago National Referral Hospital (MNRH), Butabika National Mental Hospital (BNMH), and Mbarara Regional Referral Hospital (MRRH). Outcomes from this study can inform specific targets of future interventions, and can provide insight towards improving the provision of epilepsy care in Uganda. The present thesis focuses on a narrower objective of the overall project, and involves validating outcome measures for PWE in Uganda, as well as using these measures in a preliminary analysis to identify predictors that most influence health outcomes.

Methods: The present study is nested in a larger prospective cohort study that recruits patients of all ages seeking care for epilepsy at MNRH, BNMH, and MRRH, and conducts follow-up over the phone at three weeks, three months and six months. At the hospital baseline visit, patients and/or caregivers were interviewed in either English, Luganda, or Runyankole to obtain demographic and clinical information and the Personal Impact of Epilepsy Scale (PIES), among other measures conducted as a part of a longer one- to two-hour interview. Reliability, internal consistency specifically, was assessed using three parameters: Cronbach’s alpha, McDonald’s Omega, and composite reliability for the PIES. Construct validity (internal structure) was evaluated with principal components analysis (PCA) and confirmatory factor analysis (CFA). Based on the validation results, the PIES questionnaire was treated as three outcome measures, from three subscales of the PIES. Possible predictors of the subscale scores were assessed using bivariate and multivariable linear regression.

Results: 626 patients seeking care for epilepsy and/or their caregivers were interviewed at baseline. The three-factor model of the PIES had excellent reliability. The PCA and CFA models for the scale demonstrated adequate fit with the TLI, CFI and RMSEA indices, but the model demonstrated inadequate fit with the Chi-square indicator. Several demographic and clinical indicators were associated with the three subscale scores of the PIES, Seizures, Adverse Effects of Medications, and Mood and Social Situation, in the multivariable regression models. Of these indicators, for the PIES Seizures Subscale Score, recruitment site and experience of confusion or tiredness after seizures were negatively associated with outcome scores, while taking one anti-epileptic drug (AED), education, and employment were positively associated with outcome scores. For the PIES Adverse Effects of Medications Subscale Score, taking AEDs, recruitment at MRRH, experience of isolated repetitive movements, auras, learning difficulties, and fevers were negatively associated with outcome scores, while employment status was positively associated with outcome scores. Finally, for the PIES Mood and Social Situation Subscale, recruitment at MRRH, experience of confusion and tiredness after seizures, learning difficulties, and phenytoin use were negatively associated with outcome scores, while education and employment were positively associated with outcome scores.

Conclusions: This thesis presents the first Luganda and Runyankole versions of the PIES, and the first validation of this scale with epilepsy patients in Uganda. The PIES has demonstrated adequate reliability and validity, and can be used to assess epilepsy related health outcomes, which is especially important in low-resource settings, where diagnostic equipment and specialist providers for epilepsy are not readily accessible. Furthermore, this thesis presents some preliminary results from the overall cohort study that use the PIES scale to identify some potential predictors of epilepsy-related health outcomes. Among the variables that showed associations with the PIES subscale scores, education and employment were associated with all three scores, highlighting the importance of community- and policy-based interventions for PWE, in addition to the possible identification of epilepsy subpopulations that may be at higher risk for poorer outcomes during clinical interventions.