TABLE 2

Study Quality Assessments Using the Newcastle-Ottowa Scale

VariableStudy
Ponsonby et al,16 1988–1991Fleming et al,18 1993–1996Hauck et al,17 1993–1996Blair et al,15 2003–2006
Selection
 Is the case definition adequate (with independent validation)?Yes, after postmortem examination (1*)Yes, using the Avon Clinico-Pathological Classification System (1*)Yes, with independent validation, as determined by the Office of the Medical Examiner (1*)Yes, using the Avon Clinico-Pathological Classification System (1*)
 Representativeness of the case subjectsConsecutive series using professional state ambulance service covering 94% of the population (1*)Consecutive series using established notification network system (1*)Consecutive series using Office of the Medical examiner notification system (1*)Consecutive series using established notification network system (1*)
 Selection of control subjectsCommunity control subjects, from maternity lists (1*)Community control subjects, from health visiting lists (1*)Community control subjects, from ongoing review of birth certificates (1*)Community control subjects from maternity wards, weighted for selection bias according to maternal social class (1*)
 Definition of control subjectsNo history of disease (infant has not died) (1*)No history of disease (infant has not died) (1*)No history of disease (infant has not died) (1*)No history of disease (infant has not died) (1*)
Comparability
 Comparability of case subjects and control subjects on the basis of the design or analysis(1) Study controls for infant age (1*); (2) Study controls for additional factors (time of last or reference sleep) (1*)(1) Study controls for infant age (1*); (2) Study controls for additional factors (time of last or reference sleep) (1*)(1) Study controls for infant age (1*); (2) Study controls for additional factors (maternal ethnicity, infant birth weight, time of last or reference sleep) (1*)(1) Study controls for infant age (1*); (2) Study controls for additional factors (random control subjects matched for social class) (1*)
Exposure
 Ascertainment of exposureInterviewer not blinded to case/control statusInterviewer not blinded to case/control statusInterviewer not blinded to case/control statusInterviewer not blinded to case/control status
 Same method of ascertainment for case infants and control subjectsYes, semi-structured interviews (1*)Yes, semi-structured interviews (1*)Yes, semi-structured interviews (1*)Yes, semi-structured interviews (1*)
 Nonresponse rateNonrespondents described: 6% case subjects, 17% control subjectsSame rate for both groups: 8.7% case subjects, 7.5% control subjects (1*)Nonrespondents described: 0.0% case subjects, 95.5% control subjectsSame rate for both groups: 4% case subjects, 5% random control subjects (1*)
Total stars (of 9)7/98/97/98/9