Background
Methods
Search strategy
Screening process
Criteria | Inclusion | Exclusion |
---|---|---|
Model type | Prognostic models | Diagnostic models* |
Intended scope of the review | Inform decision making at individual level (e.g. using risk of in-hospital mortality) and planning service delivery (e.g. nursing staffing) | |
Types of modelling studies | Development and/or validation | |
Target population | Neonates† admitted to a neonatal unit in any country | Studies limited to neonates with congenital anomalies older children or adults |
Predictors | Any use of treatments or interventions as predictors | Non-therapeutic intervention, e.g. radiological imaging intensity Treatment intensity not measured by enumeration of therapeutic interventions |
Outcomes | Any outcome | |
Time of prediction | No restriction | |
Intended moment of use | No restriction |
Data extraction and critical appraisal of individual studies
Descriptive analyses
Certainty of the evidence across studies
Results
Study selection
Study characteristics
Study | Study dates | Objective* | Setting | Sample size | In-hospital mortality||
|
---|---|---|---|---|---|
Georgieff, 1989 [40] | 1987 | Validation (TISS)‡
| 1 NICU†, USA | 55 | 0 |
Gray, 1992 [19] | 1989–1990 | Development (NTISS)§
| 3 NICUs†, USA | 1768 | 114 |
Davies, 1995 [33] | Not reported | Validation (NTISS)§
| 1 NICU†, South Africa | 50 | 8 |
Eriksson, 2002 [34] | 1991–1995 | Validation (NTISS)§
| 2 NICUs†, Sweden | 240 | 39 |
Zupancic, 2002 [35] | 1998 & 1999 | Validation (NTISS)§
| 1 NICU†, USA | 154 | Not reported |
Mendes, 2006 [35] | 2004 | Validation (NTISS)§
| 2 NICUs†, Brazil | 96 | 9 |
Rojas, 2011 [37] | 2007 | Validation (NTISS)§
| 1 NICU†, 1 intermediate unit, Colombia | 22 | Not reported |
Oygur, 2012 [38] | 2006–2010 | Validation (NTISS)§
| 1 NICU†, Turkey | 364 | 103 |
Shah, 2015 [32] | 2010–2012 | Development (unnamed) | 23 NICU†s, Canada | 9978 | 650 |
Wu,2015 [39] | 2007–2011 | Validation (NTISS)§
| 1 NICU†, Taiwan | 172 | 18 |
Critical appraisal of study design and statistical analysis of individual studies
All studies
Study | Study type* | Participants†
| Outcome(s)‡
| Predictors§
| Sample size||
| Missing data¶
|
---|---|---|---|---|---|---|
Georgieff,1989 [40] | - | - | + | + | + | ? |
Gray,1992 [19] | - | - | + | + | ? | + |
Davies,1995 [33] | + | + | + | + | + | ? |
Eriksson,2002 [34] | - | - | + | + | + | + |
Zupancic,2002 [35] | - | + | ? | + | + | ? |
Mendes,2006 [36] | - | - | + | + | + | + |
Rojas,2011 [37] | - | - | + | + | + | ? |
Oygur,2012 [38] | + | - | + | + | + | - |
Shah,2015 [32] | + | - | + | + | ? | + |
Wu,2015 [39] | + | - | + | + | + | ? |
Study | Study type | Performance†
| Validation‡
| Presentation§
|
---|---|---|---|---|
Georgieff,1989 [40] | Evaluation of existing score* | ? | ? | ? |
Gray, 1992 [19] | Development of new score | + | ? | ? |
Davies, 1995 [33] | Evaluation of existing score* | ? | ? | ? |
Eriksson, 2002 [34] | Evaluation of existing score* | + | ? | ? |
Zupancic, 2002 [35] | Evaluation of existing score* | ? | ? | ? |
Mendes, 2006 [36] | Evaluation of existing score* | ? | ? | ? |
Rojas, 2011 [37] | Evaluation of existing score* | ? | ? | ? |
Oygur,2012 [38] | Evaluation of existing score* | + | ? | ? |
Shah et al., 2015 [32] | Development of new score | + | ? | ? |
Wu, 2015 [39] | Evaluation of existing score* | + | ? | ? |
Score development studies
Application of existing scores to external populations
Synthesis of results of estimated performance across studies
Performance with respect to mortality
Outcome | No. of studies | Factors that may decrease certainty of evidence* | Overall† certainty | Importance ‡ | ||||
---|---|---|---|---|---|---|---|---|
Limitations | Indirectness | Inconsistency | Imprecision | Reporting bias | ||||
Mortality | 5 studies (n = 2594) | Serious | Serious | None | None | Unlikely | ⊕⊕OO low | Critical |
Morbidity | 2 studies (n = 295) | Serious | Serious | None | None | Unlikely | ⊕⊕OO low | Critical |
Composite (Morbidity and Mortality) | 2 studies (n = 10,218) | Serious | Serious | None | None | Unlikely | ⊕⊕OO low | Important |
Nursing workload | 3 studies (n = 317) | Serious | Serious | None | None | Unlikely | ⊕⊕OO low | Critical |
Hospital Costs | 1 study (n = 1768) | Serious | Serious | None | None | Unlikely | ⊕⊕OO low | Important |
Length of stay | 1 study (n = 1768) | Serious | Serious | None | None | Unlikely | ⊕⊕OO low | Important |
Time inputs | 1 study (n = 154) | Serious | Serious | None | None | Unlikely | ⊕⊕OO low | Critical |
Comparison of resource use | 1 study (n = 96) | Serious | Serious | None | None | Unlikely | ⊕⊕OO low | Important |
Rehabilitation | 1 study (n = 240) | Serious | Serious | None | None | Unlikely | ⊕⊕OO low | Not important |