Quality improvement in parent-child reviews

Article type
Authors
Hayward S, Brunton G, Thomas H, Ciliska D
Abstract
Objective: To assess any difference in the qualify of retrieved reviews for a systematic overview evaluating the effectiveness of parent-child interventions in public health, from time 1 (1985 - 1993) to time 2 (1994-1996).

Methods: Using a devised strategy, review articles within the scope of Canadian Public Health Nursing practice were retrieved from online and CD-ROM MEDLINE and CINAHL databases, hand searched key journal articles and bibliographic references during initial retrieval 1979 - 1993 (time 1), and an update 1994 - 1996 (time 2). Articles were evaluated for relevance and validity using pretested and revised tools. Two reviewers read a subset of articles, achieving interrater reliability (kappa = 0.8). One reviewer rated the remaining articles. Articles were reviewed for qualify based on a) the search strategy description, b) the search strategy comprehensiveness, c) the description of primary article selection, d) the description of primary study qualify, e) the comprehensiveness of primary study qualify, f) the findings explained considering primary study qualify, g) the methods of combining findings reported, h) appropriate analysis of findings and i) adequacy of data to support conclusions. Articles meeting eight or nine of the above criteria were graded of strong quality, those meeting six or seven were of moderate quality, and those meeting 5 or less criteria points were of weak quality.

Results: At time 1, a total of 71 articles were retrieved and 36 were relevant. Of those relevant articles, nine were of strong quality, eleven were of moderate quality, and sixteen were of weak quality. At time 2, 51 articles were retrieved and 27 were relevant. Quality ratings were as follows: sixteen strong, five moderate and six weak. Most improved rating from time 1 to time 2 was in the reporting of methods of combining findings, at a 30% increase. Least improved rating from time 1 to time 2 was a 6% decrease in the comprehensiveness of quality assessment of reviews.

Discussion: Overall, there has been an increase in the number of reviews of parent-child health interventions published. The quality of these reviews has improved. Comprehensive assessment using criteria must be included in reporting review results.