Table 1.
Included studies
| Study | Country | Hospitals | Unit type(s) | Participants | n | Missed care measure | Validityb | Effect of nurse staffingc(summary) | Main results | |
|---|---|---|---|---|---|---|---|---|---|---|
| Internal | External | |||||||||
| Al‐Kandari and Thomas (2009) | Kuwait | 5 | Med/surg | RN | 780 | IHOC survey | − | + | + | Significant positive correlation between the number of RNs in unit and adequate documentation of nursing care (p < .005) |
| Ausserhofer, Zander, Busse, Schubert, De Geest, Rafferty, and Consortium (2014)a | Europe | 488 | Med/surg | RN | 33659 | RN4CAST | + | ++ | + | Fewer nursing care tasks left undone in hospitals with lower p:n ratios and more favourable work environment (β = 0.09; p<.0001). |
| Ball, et al. (2014)a | England | 46 | Med/surg | RN | 2,917 | RN4CAST | + | ++ | + | Fewer patients per nurse fewer elements of care missed (p < .001; OR = 0.34, 95% CI 0.22–0.53). |
| Ball, et al. (2016)a | Sweden | 79 | Med/surg | RN | 10,174 | RN4CAST | + | ++ | + | Fewer than four patients per RN reduced the odds of missed care by 85% (OR 0·148, p < ·001). Compared with shifts of 10+ patients per RN, odds of missed care more than halved on shifts of six or fewer patients per RN (p < ·001, OR 0·466). |
| Bruyneel, et al. (2015)a | Europe | 217 | Med/surg | RN | 10733 | RN4CAST | + | ++ | + | Fewer clinical care tasks left undone (hypothesized hospital‐level mediator) was significantly associated with lower p:n ratio 0.039 CI (credibility intervals) [0.002, 0.072]. |
| Cho, Kim, Yeon, You, and Lee (2015) | Korea | 1 | General hospital (NS) | RN | 232 | MISSCARE | − | − | + | High staffed units (patient per nurse ratio 7:1) were associated with less missed nursing care than low staffed units (patient per nurse ratio 17:1) (β 0.136; p = .02) |
| Cho, Lee, Kim, Kim, Lee, Park, and Sung (2016) | Korea | 51 | General hospital (all) | RN | 3,037 | RN4CAST | + | + | + | One additional patient per RN was associated with a 3% increase in the odds of care left undone (OR = 1.03 p < .001) |
| Dabney & Kalisch (2015) | USA | 2 | General hospital (NS) | Patient | 729 | MISSCARE patient | − | − | + | Patients’ reports of not receiving timely care weakly significantly correlated with RN hours per patient day (r = −.14, p = .002) |
| Friese, Kalisch, and Lee (2013) | USA | 9 | Oncology | RN & HCSW | 2,318 | MISSCARE | − | − | + | Increase in the number of patients cared for by RN and HCSW was associated with a 2.1% increase in the total missed nursing care score (p < .05) |
| Griffiths, Dall'Ora, et al., 2014, a | Europe | 488 | Med/surg | RN | 31,627 | RN4CAST | + | ++ | + | Odds of nurses leaving care undone increased by 26% when nurses were caring for >11.5 patients, compared with nurses caring for ≤6 patients (OR = 1.26; 95% CI = 1.23–1.29) |
| Kalisch, Tschannen, Lee, and Friese, 2011 | USA | 10 | General hospital (all) | RN & HCSW | 4,086 | MISSCARE | + | + | + | Staff that cared for more patients reported more missed care (β = 0.015 p < .001) |
| Kalisch, Tschannen, and Lee, 2011 | USA | 10 | General hospital (all) | RN & HCSW | 4,288 | MISSCARE | + | + | + | Higher nursing hours per patient day were associated with lower level of missed care (β = −0.45 p = .002) |
| Kalisch, Doumit, Lee, and Zein (2013)c | USA/Lebanon | 2 | Med/surg/ICU | RN |
633 114 |
MISSCARE | − | − | (+) | Number of patients cared for was not a significant predictor of missed care |
| Orique, Patty, Woods (2015) | USA | 1 | General hospital (all) | RN & HCSW | 169 | MISSCARE | − | − | + | No significant relationship between unit‐level nurse workload and missed nursing care but significant positive relationship between missed care and number of patients under care (r = .246, p = .001) |
| Palese, Ambrosi, Prosperi, Guarnier, Barelli, Zambiasi, and Saiani (2015) | Italy | 12 | Med | RN & HCSW |
205 109 |
MISSCARE | + | + | + | Lower p:n was associated with less missed care (OR = 0.91 p < .05) |
| Schubert, Ausserhofer, Desmedt, Schwendimann, Lesaffre, Li, and De Geest (2013)a | Swiss | 35 | Med/surg | RN | 1,633 | BERNCA revised | + | ++ | + (+) | Shifts with ≤6 patients per RN were associated with a 53% reduction in the odds of care left undone, compared with shifts on which there were ≥10 patients per RN (OR = 0.47, p < .001). |
| Weiss, Yakusheva, and Bobay (2011) | USA | 4 | Med/surg | Patient | 1,892 | QTDS (delivered) | + | + | (−) | No significant association between non‐overtime RNHPPD and patient reported delivery of necessary discharge information (β −0.05 p = .74). |
| Zander, Dobler, Baumler, and Busse (2014)a | German | 49 | Med/surg | RN | 1,511 | RN4CAST | + | ++ | + (+/−)− | Missing patient surveillance, skincare and medication given on time were increased by 3% for each additional patient per nurse (OR = 1.03 p < .01). Some aspects of care showed smaller but significant associations in the opposite direction; no overall relationship. |
a
Studies that are based on the data collected as part of the RN4Cast study in 12 European countries Belgium, England, Finland, Germany, Greece, Ireland, The Netherlands, Norway, Poland, Spain, Sweden and Switzerland. Where “Europe” is indicated the study analysis was conducted by pooling data across some or all of these countries.
b
− High risk of bias—few criteria fulfilled conclusions likely to be affected. + moderate risk of bias—several criteria fulfilled conclusions unlikely to be affected. ++ low risk of bias = most criteria fulfilled conclusions very unlikely to be affected.
c
+ Statistically significant association showing a benefit (reduced missed care from higher staffing levels). − statistically significant effect showing harm (increased missed care with higher staffing levels). (+) non‐significant association showing direction of effect or no effect.
NS, specialities not specified. p:n patients per nursing staff (RN only or RN + HCSW).