Validity of the Personal Impact of Epilepsy Scale (PIES) in patients with epilepsy in Uganda

Epilepsy Behav. 2021 Jan;114(Pt B):107303. doi: 10.1016/j.yebeh.2020.107303. Epub 2020 Jul 25.


Purpose: We sought to address the construct validity and reliability of the Personal Impact of Epilepsy Scale (PIES), an epilepsy-specific quality-of-life measure, in patients with epilepsy in Uganda. We also sought to assess the applicability of the scale across three languages: English, Luganda, and Runyankole.

Methods: Patients with epilepsy (N = 626) were recruited at the time of care seeking from Mulago National Referral Hospital (MNRH), Butabika National Referral Mental Hospital (BNRMH), and Mbarara Regional Referral Hospital (MRRH), and were given the English, Runyankole, and Luganda versions of the PIES as a part of a larger interview. Reliability, internal consistency specifically, was assessed using three parameters: Cronbach's alpha, McDonald's Omega, and composite reliability. Construct validity (internal structure) was evaluated with principal component analysis (PCA) for three factors, as well as confirmatory factor analysis (CFA) for a three-factor structure of the scale. We also assessed correlations between the three PIES subscales and the seizure severity question in the Liverpool Seizure Severity Scale (LSSS) and reported seizure frequency.

Results: The three-factor model of the PIES had adequate reliability, with Cronbach's Alpha, McDonald's Omega, and composite reliability values over 0.7, except for the Cronbach's Alpha and McDonald's Omega values for the second factor, which was slightly lower than 0.7 in the full sample as well as when stratified by study language. The PCA and CFA models for the scale demonstrated adequate fit with the Tucker-Lewis index (TLI), comparative fit index (CFI), and root mean square error of approximation (RMSEA), with TLI and CFI values above 0.9 and RMSEA values less than 0.08. However, the model demonstrated inadequate fit with the Chi-square indicator, which yielded a significant p-value. Individual factor loadings ranged from 0.50 to 0.95 in the full sample, 0.45 to 0.98 in the English sample, and 0.45 to 0.93 in the Luganda sample. Finally, the three PIES subscales aligned with reported seizure frequency and the seizure severity question from the LSSS.

Conclusions: This study presents the first Luganda and Runyankole versions of the PIES, and the first validation of this scale in English and Luganda with patients with epilepsy in Uganda. The PIES was found to have acceptable psychometric properties for reliability and validity parameters. Thus, the scale is recommended for use and for further investigation in patients with epilepsy in Uganda. This article is part of the Special Issue "The Intersection of Culture, Resources, and Disease: Epilepsy Care in Uganda".

Keywords: Epilepsy; Quality of life; Scales; Uganda; Validation.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Epilepsy* / diagnosis
  • Humans
  • Psychometrics
  • Reproducibility of Results
  • Surveys and Questionnaires
  • Uganda