Conference abstract
Missing data in the charts of HIV-positive clients on antiretroviral treatment and its handling in a retrospective cross-sectional study in the West Region of Cameroon
Pan African Medical Journal - Conference Proceedings. 2023:18(161).03
Oct 2023.
doi: 10.11604/pamj-cp.2023.18.161.2278
Archived on: 03 Oct 2023
Contact the corresponding author
Keywords: Data, medical records, patient, health
Oral presentation
Missing data in the charts of HIV-positive clients on antiretroviral treatment and its handling in a retrospective cross-sectional study in the West Region of Cameroon
Djouma Nembot Fabrice1,&, Emmanuel Nshom2, Emile Tata2, Larry Ngassa2, Ghislaine Feugo2, Ismaila Esa2, Mboh Eveline Asongwe2, Jerome Ateudjieu1
1Department of Public Health University of Dschang, Dschang, Cameroon, 2Cameroon Baptist Convention Health Services, Cameroon
&Corresponding author
Introduction: good keeping of medical records is still challenging in resource-limited settings where paper-based documentation remains predominant. Missing data in a patient’s chart harms patients and health service management and research that uses secondary data, especially in the context of increased need for evidence. This study reports the rate of missing information in Charts of patients on Antiretroviral Treatment (ART) and how it has been handled in a retrospective cross-sectional study done in the West Region of Cameroon.
Methods: it was a cross-sectional study. Patient charts of clients initiated on ART from October 2019 to September 2020 in 25 high-volume clinics in the West Region of Cameroon were reviewed and 41 purposively sampled variables were reviewed to determine the completeness rate. Descriptive analysis was done by estimating the rate of missing information across health facility (HF) characteristics. The two-way ANOVA was conducted to examine the effect of facility volume and facility sector on the mean missingness rate. The methods of handling missing data in the retrospective cross-sectional study done on the data collected in the medical charts were also reported.
Results: a total of 2735 medical charts were included in the analysis. The overall mean missingness rate was 10.9% (SD: 14.3%) and 36.1% (SD: 31.9%) respectively in adults’ and infants’ charts. The two-way ANOVA test done on the adults’ charts showed a significant interaction between the effects of the facility sector and facility volume on the overall mean missingness rate, (F (1, 2595) = 12.37, p < .0001). The simple main effects analysis showed that in the private sector, high-volume HFs had a mean missingness rate significantly higher than low-volume HFs (p < .0001). In the retrospective cross-sectional study done on the data collected in the medical charts, the pairwise deletion method was used for participants with missing data on the dependent variable; explanatory variables > 25% missingness were excluded from models and the missing-indicator method was used for explanatory variables with ≤ 25% missingness.
Conclusion: missingness rate was high in the HIV-positive patients’ chart in the West Region of Cameroon, specifically in public and high-volume HFs. Interventions should be implemented to reduce the level of missing data in charts which will have a positive impact on patient’s health and services management and on the quality of the results generated by studies that use medical records as data sources.
Missing data in the charts of HIV-positive clients on antiretroviral treatment and its handling in a retrospective cross-sectional study in the West Region of Cameroon
Djouma Nembot Fabrice1,&, Emmanuel Nshom2, Emile Tata2, Larry Ngassa2, Ghislaine Feugo2, Ismaila Esa2, Mboh Eveline Asongwe2, Jerome Ateudjieu1
1Department of Public Health University of Dschang, Dschang, Cameroon, 2Cameroon Baptist Convention Health Services, Cameroon
&Corresponding author
Introduction: good keeping of medical records is still challenging in resource-limited settings where paper-based documentation remains predominant. Missing data in a patient’s chart harms patients and health service management and research that uses secondary data, especially in the context of increased need for evidence. This study reports the rate of missing information in Charts of patients on Antiretroviral Treatment (ART) and how it has been handled in a retrospective cross-sectional study done in the West Region of Cameroon.
Methods: it was a cross-sectional study. Patient charts of clients initiated on ART from October 2019 to September 2020 in 25 high-volume clinics in the West Region of Cameroon were reviewed and 41 purposively sampled variables were reviewed to determine the completeness rate. Descriptive analysis was done by estimating the rate of missing information across health facility (HF) characteristics. The two-way ANOVA was conducted to examine the effect of facility volume and facility sector on the mean missingness rate. The methods of handling missing data in the retrospective cross-sectional study done on the data collected in the medical charts were also reported.
Results: a total of 2735 medical charts were included in the analysis. The overall mean missingness rate was 10.9% (SD: 14.3%) and 36.1% (SD: 31.9%) respectively in adults’ and infants’ charts. The two-way ANOVA test done on the adults’ charts showed a significant interaction between the effects of the facility sector and facility volume on the overall mean missingness rate, (F (1, 2595) = 12.37, p < .0001). The simple main effects analysis showed that in the private sector, high-volume HFs had a mean missingness rate significantly higher than low-volume HFs (p < .0001). In the retrospective cross-sectional study done on the data collected in the medical charts, the pairwise deletion method was used for participants with missing data on the dependent variable; explanatory variables > 25% missingness were excluded from models and the missing-indicator method was used for explanatory variables with ≤ 25% missingness.
Conclusion: missingness rate was high in the HIV-positive patients’ chart in the West Region of Cameroon, specifically in public and high-volume HFs. Interventions should be implemented to reduce the level of missing data in charts which will have a positive impact on patient’s health and services management and on the quality of the results generated by studies that use medical records as data sources.
