TY - JOUR
T1 - Methods in pharmacoepidemiology
T2 - A review of statistical analyses and data reporting in pediatric drug utilization studies
AU - Sequi, Marco
AU - Campi, Rita
AU - Clavenna, Antonio
AU - Bonati, Maurizio
PY - 2013/3
Y1 - 2013/3
N2 - Purpose: To evaluate the quality of data reporting and statistical methods performed in drug utilization studies in the pediatric population. Methods: Drug utilization studies evaluating all drug prescriptions to children and adolescents published between January 1994 and December 2011 were retrieved and analyzed. For each study, information on measures of exposure/consumption, the covariates considered, descriptive and inferential analyses, statistical tests, and methods of data reporting was extracted. An overall quality score was created for each study using a 12-item checklist that took into account the presence of outcome measures, covariates of measures, descriptive measures, statistical tests, and graphical representation. Results: A total of 22 studies were reviewed and analyzed. Of these, 20 studies reported at least one descriptive measure. The mean was the most commonly used measure (18 studies), but only five of these also reported the standard deviation. Statistical analyses were performed in 12 studies, with the chi-square test being the most commonly performed test. Graphs were presented in 14 papers. Sixteen papers reported the number of drug prescriptions and/or packages, and ten reported the prevalence of the drug prescription. The mean quality score was 8 (median 9). Only seven of the 22 studies received a score of ≥10, while four studies received a score of
AB - Purpose: To evaluate the quality of data reporting and statistical methods performed in drug utilization studies in the pediatric population. Methods: Drug utilization studies evaluating all drug prescriptions to children and adolescents published between January 1994 and December 2011 were retrieved and analyzed. For each study, information on measures of exposure/consumption, the covariates considered, descriptive and inferential analyses, statistical tests, and methods of data reporting was extracted. An overall quality score was created for each study using a 12-item checklist that took into account the presence of outcome measures, covariates of measures, descriptive measures, statistical tests, and graphical representation. Results: A total of 22 studies were reviewed and analyzed. Of these, 20 studies reported at least one descriptive measure. The mean was the most commonly used measure (18 studies), but only five of these also reported the standard deviation. Statistical analyses were performed in 12 studies, with the chi-square test being the most commonly performed test. Graphs were presented in 14 papers. Sixteen papers reported the number of drug prescriptions and/or packages, and ten reported the prevalence of the drug prescription. The mean quality score was 8 (median 9). Only seven of the 22 studies received a score of ≥10, while four studies received a score of
KW - Adolescent
KW - Child
KW - Drug utilization
KW - Pharmacoepidemiology
KW - Review
UR - http://www.scopus.com/inward/record.url?scp=84877118925&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84877118925&partnerID=8YFLogxK
U2 - 10.1007/s00228-012-1354-y
DO - 10.1007/s00228-012-1354-y
M3 - Article
C2 - 22832725
AN - SCOPUS:84877118925
VL - 69
SP - 599
EP - 604
JO - European Journal of Clinical Pharmacology
JF - European Journal of Clinical Pharmacology
SN - 0031-6970
IS - 3
ER -