The representation of vulnerable populations in quality improvement studies.

PURPOSE: A mapping review to quantify representation of vulnerable populations, who suffer from disparity and often inequitable healthcare, in quality improvement (QI) research. DATA SOURCES: Studies published in 2004-2014 inclusive from Medline, Embase and Cochrane databases for English language research with the terms 'quality improvement' or 'quality control' or 'QI' and 'plan-do-study-act' or 'PDSA' in the years 2004-2014 inclusively. STUDY SELECTION: Published clinical research that was a QI-themed, as identified by its declared search terms, MESH terms, abstract or title. DATA EXTRACTION: Three reviewers identified the eligible studies independently. Excluded were publications that were not trials, evaluations or analyses. RESULTS OF DATA SYNTHESIS: Of 2039 results, 1660 were eligible for inclusion. There were 586 (33.5%) publications that targeted a specific vulnerable population: children (184, 10.54%), mental health patients (125, 7.16%), the elderly (100, 5.73%), women (57, 3.27%), the poor (30, 1.72%), rural residents (29, 1.66%), visible minorities (27, 1.55%), the terminally ill (17, 0.97%), adolescents (16, 0.92%) and prisoners (1 study). Seventy-four articles targeted two or more vulnerable populations, and 11 targeted three population categories. On average, there were 158 QI research studies published per year, increasing from 69 in 2004 to 396 in 2014 (R2 = 0.7, P < 0.001). The relative representation of vulnerable populations had a mean of 33.58% and was stable over the time period (standard deviation (SD) = 5.9%, R2 = 0.001). Seven countries contributed to over 85% of the publications targeting vulnerable populations, with the USA contributing 62% of the studies. CONCLUSIONS: Over 11 years, there has been a marked increase in QI publications. Roughly one-third of all published QI research is on vulnerable populations, a stable proportion over time. Nevertheless, some vulnerable populations are under-represented. Increased education, resources and attention are encouraged to improve the health of vulnerable populations through focused QI initiatives.

Nil-per-os days and necrotizing enterocolitis in extremely preterm infants.

OBJECTIVE: This study aims to evaluate the association between nil-per-os (NPO) days and development of necrotizing enterocolitis (NEC) in extremely preterm neonates (<29 weeks gestational age). STUDY DESIGN: A case-control study of 234 extremely preterm neonates who developed stage II or III NEC and 467 matched control infants admitted to participating sites in the Canadian Neonatal Network between 2010 and 2011 was conducted. The number and percentage of NPO days before the development of NEC was compared with the equivalent period in control infants using logistic regression. RESULTS: Infants with NEC were NPO on average 5.6 days (28% of days before NEC) versus 3.7 days (19% of equivalent days; p < 0.01) among controls. NEC cases required more days of inotropic support, antibiotic use, and had higher rates of patent ductus arteriosus (PDA). After adjusting for inotrope use, PDA, antibiotics, and severity of illness, for each additional NPO day the adjusted odds for NEC increased by 1.08 (95% confidence interval: 1.04-1.12). CONCLUSION: Among extremely preterm neonates, those who developed NEC were NPO for a longer period of time than control infants who were NEC-free. We speculate that delayed initiation or interruption of feeding may be a potent risk factor for NEC.