Decreased Inaccuracies in HIV Screening following Strengthening of Quality System: Evidence-based Interventions through the PMTCT Program in Cameroon

ABSTRACT Objective. To evaluate the evolution in error rates of HIV testing and the quality of HIV testing within the national PMTCT program before and after quality assurance interventions. Methods. We conducted a study from 2012 to 2016 among pregnant women in Cameroon. The quality system was evaluated with a checklist. Performance in HIV screening (sensitivity, specificity, positive predictive and negative values) was evaluated using the national reference laboratory as gold standard. Based on quality performance, a package of quality interventions was implemented and evaluated thereafter. Results. In 2012 and 2016, 6521 (93.2%) and 6859 (97.99%) eligible pregnant women were enrolled. Following quality interventions between the two surveys, a significant increased performance was reported from 2012 (average score: 56.25%) to 2016 (average score: 79.75%); p=0.0157. Sensitivity in HIV testing increased from 81.2% to91.4% (p<0.00001); specificity increased from 99.3% to 99.6% (p=0.0058); positive- and negative-predictive values also increased respectively: 90.2% and 98.7% (2012) to 93.5% and 99.5% (2016); p<0.001. Conclusion. Evidence-based quality interventions are effective in reducing errors occurred during routine HIV testing. However, increased performance merits further efforts to meet the required standards (97.5% sensitivity and 99.8% specificity). Thus, implementing periodic quality interventions and re-testing of HIV-positive specimens will reduce diagnostic errors. RESUME Objectif. Evaluer la variation des taux d’erreur dans les tests de VIH et le systeme qualite dans le programme national de PTME avant et apres des interventions programmatiques sur l’assurance qualite. Methodes. Une etude a ete menee de 2012 a 2016 chez les femmes enceintes au Cameroun. A l’aide d’un checklist, le systeme qualite a ete evaluate. La performance diagnostic du VIH (sensitivite, specificite, valeurs predictives positive et negative) etait evaluatee les rersultats du laboratoire de reference comme standard. Sur la base de la performance, un paquet d’interventions qualite a ete mis en œuvre et son impact evalue entre 2012 et 2016. Resultats. En 2012 et 2016, 6521 (93,2%) et 6859 (97,99%) de femmes enceintes eligibles etaient enrolees. Suivant les interventions qualite mises en oeuvre, une amelioration significative de la performance generale a ete observee de 2012 (score moyen: 56,25%) a 2016 (score moyen: 79,75%); p=0,0157. La sensibilite du HIV s’est amelioree significativement de 81,2% a 91,4% (p<0,00001); la specificite s’est amelioree significativement de 99,3% a 99,6% (p=0,0058); les valeurs predictives positive et negative sont passees respectivement de 90,2% et 98,7% (2012) a 93,5% et 99,5% (2016) ; p<0,001. Conclusion. Les interventions probantes sur le systeme qualite se traduisent par baisse significative des erreurs dans le test du VIH en routine. Toutefois, de efforts sont necessaires pour optimiser les performance vers les seuils standards (sensibilte : 97,5% ; specificite : 99,8%). Instituer une periodicite des interventions qualite et le re-testing des echantillons positifs permettrait de baisser davantage ces erreurs.