Background
Methods
Eligibility criteria, information sources, search strategy
Inclusion criteria | Exclusion criteria | |
---|---|---|
POPULATION | all births, term/ normal birth weight birth or low risk birth in a nationwide setting with < 5/1000 neonatal deaths | Preterm birth, low birth weight birth, other risk-selections (e.g. gestational diabetes, multiple births) |
EXPOSITION | comparison of different hospital birth volumes or -sizes | No comparison of different hospital birth volumes or -sizes |
COMPARISON | other birth volumes | No comparator provided |
OUTCOME | Primary Outcome: Maternal or infant mortality Secondary Outcomes: Caesarean sections, readmissions, birth complications, developmental delays | No measurement of maternal or infant mortality |
STUDY TYPE | Observational and interventional studies | Descriptive studies, systematic reviews |
Study selection
Data extraction and data synthesis
Critical appraisal process
Patient and public involvement
Results
Study selection
Study characteristics
Study | Period | Country | Birth population | Grouped annual hospital volume | Outcomes | Outcome definition |
---|---|---|---|---|---|---|
Finnstrom et al. 2006[27] | 1985–1999 | SWE | births: all singletons (n = 1.538.814) | < 500, 500–999, 1000–2499 (ref.), ≥ 2500 | 1) neonatal mortality | 1) ≤ 27d |
Friedman et al. 2016[28] | 1998–2010 | US | women: all hospital (n = 50.433.539) | 50, 1000 (ref.), 1500, 2250 | 1) maternal mortality 2) maternal complications | 1) failure to rescue 2) severe morbidity1 |
Heller et al. 2002[29] | 1990–1999 | GER | births: BW > 2500 g (n = 582.655); | ≤ 500, 501–1000, 1001–1500, > 1500 (ref.) | 1) Early-neonatal death | 1) ≤ 7d |
Hemminki et al. 2011[30] | 1991–2008 | FIN | births: all (n = 474.419) + BW > 2499 g in non-UH | < 750, 750–1499, ≥ 1500, UH (ref.) | 1) perinatal mortality 2) CS | 1) ≤ 7d |
Joyce et al. 2004[31] | 1994–1996 | UK | births: all (n = 540.834) | N/A: Volume entered the analysis as continuous variable | 1) stand. stillbirth rates 2) stand. neonatal mortality | 1) > 24 wk GA 2) ≤ 28d |
Karalis et al. 2016[32] | 2005–2009 | FIN | births: low risk2 (n = 276.066) | births: ≤ 999, 1000–1999, ≥ 2000, UH (ref.) | 1) stillbirths 2) early neonatal death | 1) Intrapartum: undefined 2) undefined |
Moster et al. 2001[33] | 1967–1996 | NO | births: all (n = 1.650.852) | ≤ 100, 101–500, 501–1000, 1001–2000, 2001–3000, > 3000 (ref.) | 1) neonatal mortality | 1) ≤ 28d |
Pyykonen et al. 2014[34] | 2006–2010 | FIN | women: all3 (n = 290.288) + low risk4 (n = 276.287) | < 1000, 1000–2999 (ref.), < 3000 | 1) perinatal mortality 2) neonatal mortality 3) early neonatal mortality 4) stillbirths | 1) stillbirth + death ≤ 7d 2) ≤ 28d 3) ≤ 7d 4) ≥ 22wk GA |
Snowden et al. 2012[35] | 2006 | US | women: all (n = 527.617), low risk5 | Urban: ≤ 50–1199 (ref.), 1200–2399, 2400–3599; ≥ 3600 Rural: 50–599 (ref.) 600–1699; ≥ 1700 | 1) neonatal mortality | 1) undefined |
Tracy et al. 2006[36] | 1999–2001 | AUS | births: low risk/ term6 (n = 331.147) | < 100, 100–500, 501–1000, 1001–2000, > 2001 (ref.) | 1) neonatal mortality 2) CS (labour) 3) Overall CS | 1) ≤ 28d |
de Graaf et al. 2010[38] | 2000–2006 | NEL | women: singleton (n = 655.961) | < 750, 750–999, 1000–1249, 1250–1499, 1500–1749, ≥ 1750 (ref.) | 1) perinatal mortality 2) neonatal complications | 1) ≤ 7d 2) Perinatal adverse outcome7 |
Restrepo et al. 2018[39] | 2012 | US | births: live 20–44 wk GA (n = 32.140) | N/A: Volume entered the analysis as continu-ous variable | 1) neonatal mortality | 1) ≤ 28d |
Aubrey-Brassler et al. 2019[37] | 2006–2009 | CA | women: all (n = 820.761)/ births: all (n = 827.504) | No services usually; 1–49; 50–99; 100–199; 200–499; 500–999; 1000–2499, > 2500 (ref.) | 1) perinatal mortality 2) maternal complications | 1) Death […]8 2) Maternal Morbidity & Mortality9 |
Results of the critical appraisal
Item | Description | Finnstrom et al. 2006 [27] | Friedman et al. 2016 [28] | Heller et al. 2002 [29] | Hemminki et al. 2011 [30] | Joyce et al. 2004 [31] | Karalis et al. 2017 [32] | Moster et al. 2001 [33] | Pyykonen et al. 2014 [34] | Snowden et al. 2012 [35] | Tracy et al. 2006 [36] | de Graaf et al. 2010 [38] | Restrepo et al. 2018 [39] | Aubrey-brassler et al. 2019[37] |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1.1 | appropriate and clearly focused question | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
1.2 | illustrated comparability between studied groups | Yes | Yes | Yes | Yes | Yes | Yes | No | Yes | Yes | Yes | Yes | Yes | Yes |
1.3 | number of asked people (prospective studies) | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
1.4 | Likelihood that some eligible subjects might have the outcome at the time of enrolment is assessed and taken into account in the analysis | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
1.5 | Drop-Out rate (prospective studies) | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
1.6 | Comparison between full and lost-to-follow-up participants (prospective studies) | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
1.7 | Clearly defined outcomes | Yes | Yes | Yes | No | Yes | No | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
1.8 | Assessment of outcome blinded to exposure status | No | No | No | No | No | No | No | No | No | No | No | No | No |
1.9 | When blinding impossible, recognition that knowledge of exposure status could have influenced assessment | No | No | No | No | No | No | No | No | No | No | No | No | No |
1.10 | reliable measurement of exposure | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
1.11 | from other sources is used to demonstrate that the method of outcome assessment is valid and reliable (clearly defined primary outcomes) | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
1.12 | Exposure level or prognostic factor is assessed more than once (prospective studies) | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
1.13 | confounders identifed and adequately taken into account for analysis | Yes | Yes | Yes | Yes | Yes | Yes | Yes | No | Yes | Yes | Yes | Yes | Yes |
1.14 | confidence intervals provided | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes | No | Yes | Yes | No | Yes |
2.1 | Overall rating | Acceptable | Acceptable | Acceptable | Acceptable | Acceptable | Acceptable | Unacceptable | Acceptable | Acceptable | Acceptable | Acceptable | Acceptable | Acceptable |
-
number of participants (item 1.3)
-
outcome already present before start of study (item 1.4)
-
drop-out (item 1.5)
-
comparison between full and lost to follow-up (item 1.6) and
-
multiple measured exposure levels (item 1.12).