Kelli Carneiro de Freitas Nakata1, Priscilla Perez da Silva Pereira2, Bruno Salgado Riveros3. 1. Núcleo de Avaliação de Tecnologias em Saúde - SES/MT, Brazil. Electronic address: kellinakata@hotmail.com. 2. Laboratório de pesquisa sobre Cuidados em Saúde/LAPECS Universidade Federal de Rondônia, Brazil. 3. Universidade Federal do Paraná, Brazil.
Abstract
OBJECTIVES: to evaluate the accuracy of the creatine kinase test in neonatal screening for Duchenne Muscular Dystrophy. METHODS: Investigations were carried out in six databases, gray literature and manual search. All studies and data extraction eligibility processes were carried out in paired mode. The methodological stringency of the studies was assessed using the modified QUADAS 2 tool. The performance measures of the test were calculated using the Meta-DiSc software, version 1.4. RESULTS: 11 studies were included, with a total sample of 1,416,123 newborns. The positive likelihood ratio was more than 10 in these studies. On the other hand, the negative likelihood ratio was up to 0.2 in 9/10 studies; however, when considering the upper limit of the confidence interval, only 4/10 studies showed a negative likelihood ratio less than 0.2 and in 6/11 it was greater than 0.5. Specificity was close to 1 in studies. The sensitivity was equal to or greater than 0.81 in 10/11 studies; however, when considering the reliability's minimum interval limit, only 6/10 studies showed sensitivity equal to or greater than 0.7. Ten studies were summarized on a ROC curve indicating good performance (Area under the curve = 0.9980 and Q index = 0.9846). CONCLUSIONS: the creatine kinase test showed good accuracy in screening for cases of Duchenne Muscular Dystrophy and may be a useful alternative in the early diagnosis of the disease followed by confirmatory genetic testing.
OBJECTIVES: to evaluate the accuracy of the creatine kinase test in neonatal screening for Duchenne Muscular Dystrophy. METHODS: Investigations were carried out in six databases, gray literature and manual search. All studies and data extraction eligibility processes were carried out in paired mode. The methodological stringency of the studies was assessed using the modified QUADAS 2 tool. The performance measures of the test were calculated using the Meta-DiSc software, version 1.4. RESULTS: 11 studies were included, with a total sample of 1,416,123 newborns. The positive likelihood ratio was more than 10 in these studies. On the other hand, the negative likelihood ratio was up to 0.2 in 9/10 studies; however, when considering the upper limit of the confidence interval, only 4/10 studies showed a negative likelihood ratio less than 0.2 and in 6/11 it was greater than 0.5. Specificity was close to 1 in studies. The sensitivity was equal to or greater than 0.81 in 10/11 studies; however, when considering the reliability's minimum interval limit, only 6/10 studies showed sensitivity equal to or greater than 0.7. Ten studies were summarized on a ROC curve indicating good performance (Area under the curve = 0.9980 and Q index = 0.9846). CONCLUSIONS: the creatine kinase test showed good accuracy in screening for cases of Duchenne Muscular Dystrophy and may be a useful alternative in the early diagnosis of the disease followed by confirmatory genetic testing.