Measurement Issues in Trials of Pediatric Acute Diarrheal Diseases: A Systematic Review

Department of Clinical Epidemiology and Biostatistics, McMaster University, Health Sciences Centre, Hamilton, Ontario, Canada.
PEDIATRICS (Impact Factor: 5.3). 07/2010; 126(1):e222-31. DOI: 10.1542/peds.2009-3667
Source: PubMed

ABSTRACT Worldwide, diarrheal diseases rank second among conditions that afflict children. Despite the disease burden, there is limited consensus on how to define and measure pediatric acute diarrhea in trials.
In RCTs of children involving acute diarrhea as the primary outcome, we documented (1) how acute diarrhea and its resolution were defined, (2) all primary outcomes, (3) the psychometric properties of instruments used to measure acute diarrhea and (4) the methodologic quality of included trials, as reported.
We searched CENTRAL, Embase, Global Health, and Medline from inception to February 2009. English-language RCTs of children younger than 19 years that measured acute diarrhea as a primary outcome were chosen.
We identified 138 RCTs reporting on 1 or more primary outcomes related to pediatric acute diarrhea/diseases. Included trials used 64 unique definitions of diarrhea, 69 unique definitions of diarrhea resolution, and 46 unique primary outcomes. The majority of included trials evaluated short-term clinical disease activity (incidence and duration of diarrhea), laboratory outcomes, or a composite of these end points. Thirty-two trials used instruments (eg, single and multidomain scoring systems) to support assessment of disease activity. Of these, 3 trials stated that their instrument was valid; however, none of the trials (or their citations) reported evidence of this validity. The overall methodologic quality of included trials was good.
Even in what would be considered methodologically sound clinical trials, definitions of diarrhea, primary outcomes, and instruments employed in RCTs of pediatric acute diarrhea are heterogeneous, lack evidence of validity, and focus on indices that may not be important to participants.

Download full-text


Available from: Bradley C Johnston, Jun 18, 2015
  • [Show abstract] [Hide abstract]
    ABSTRACT: Functional constipation (FC) often begins in the first year of life. Although standard definitions and criteria have been formulated to describe FC, these are rarely used in research and clinical practice. To systematically assess how definitions and outcome measures are defined in therapeutic randomized controlled trials (RCTs) of infants with FC. PubMed, EMBASE and Cochrane databases were searched. Studies were included if (1) it was a (systematic review of) therapeutic RCT; (2) children ≤4 years old; (3) they had FC; (4) a clear definition of constipation was provided; (5) were written in English. Quality was assessed using the Delphi list. 1115 articles were found; only 5 studies fulfilled the inclusion criteria. Four different definitions were used, of which only 2 used the internationally accepted Rome-III criteria. Defecation frequency was used as primary outcome in all included trials and stool consistency in 3 trials. Two trials involving infants investigated new infant formulas, whereas the third RCT evaluated the efficacy of a probiotic strain. The 2 trials including infants up to 4 years of age compared polyethylene glycol without electrolytes (PEG4000) with lactulose and milk of magnesia. All trials used non-validated parental diaries. Different definitions and outcome measures for FC in infants are used in RCTs. Disappointingly; there is a lack of well-designed therapeutic trials in infants with constipation. In order to make comparison between future trials possible, standard definitions, core outcomes and validated instruments are needed.
    Journal of Pediatric Gastroenterology and Nutrition 11/2014; 60(4). DOI:10.1097/MPG.0000000000000631 · 2.87 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Background. Benefits associated with probiotic administration to children seeking emergency department care with diarrheal disease are unknown. Methods. In this 3-site, double-blind, placebo-controlled study, children aged 4 to 48 months with gastroenteritis were randomized to receive 5 days of placebo, low-dose (4 × 10(9) colony forming units per day), or high-dose (8 × 10(9) colony forming units per day) probiotic (Lactobacillus helveticus and Lactobacillus rhamnosus) in a 2:1:1 ratio. The primary outcome was day care absenteeism. Results. The proportion of children missing a day of day care was 63% (39/62) and 61% (37/61) in the placebo and probiotic arms, respectively (95%CI -14.6% to 18.9%). The proportions experiencing unscheduled health care provider visits and intravenous fluid rehydration were 24% (15/62), 7% (4/62), and 30% (18/61), 5% (3/61) in the placebo and probiotic study arms, respectively (P = .52 and P = .65). None of the results differed based on probiotic dose. Conclusions. The probiotic evaluated did not reduce absenteeism. Since power was suboptimal, further evaluation is required. © The Author(s) 2015.
    Clinical Pediatrics 02/2015; DOI:10.1177/0009922815569200 · 1.26 Impact Factor