Article Text

Case control study
Higher levels of nurse staffing are not associated with reduced adverse events among postoperative children if surveillance levels are low
  1. Eileen Lake
  1. School of Nursing, University of Pennsylvania, Philadelphia, Pennsylvania, USA
  1. Correspondence to : Dr Eileen Lake, School of Nursing, University of Pennsylvania, 418 Curie Blvd, Philadelphia, PA 19104, USA; elake{at}

Statistics from

Commentary on: OpenUrlCrossRefPubMed

Implications for practice and research

  • The level of nurse staffing influences the priority nurses give to surveillance; nursing units should be adequately staffed to support the surveillance of all patients.

  • Managers should routinely check with staff nurses to learn which patients they are monitoring most frequently.

  • Nurses caring for postoperative patients should give extra attention to patients with comorbidities.

  • Nursing systems research should focus on the pathways through which nursing system features influence care processes and patient outcomes.


Over the past decade, considerable evidence from cross-sectional studies has established that patient outcomes are better in hospitals with better nursing features, with staffing being the dominant feature studied.1 Other prominent features have been the composition of the nursing workforce, particularly nurse education, and the nursing practice environment.2 ,3 Hospital administrators may ignore this correlational evidence. What has been lacking is evidence that nursing system features influence care delivery. Nursing surveillance has been theorised as a key causal pathway linking nursing system features to outcomes.


A cross-sectional case–control study was conducted in an academic children's hospital to link nurse staffing and surveillance to postoperative adverse events. Two event types were studied: ‘threshold’ events requiring minor intervention, such as a change in medication, and ‘rescue’ events requiring major intervention, such as airway management or transfer to critical care. The patient electronic health records were the data source for nurse assessments and patient variables. Patients who suffered adverse events were identified by trigger tools. Two control patients were matched to each event case on surgical procedure. In sequential regression models, hypotheses were tested regarding the moderating or mediating roles of staffing and surveillance. A major strength of this study was the quasi-experimental design that increased causal inference over the traditional correlational design. The staffing measures differed for the adverse event and control groups (ie, event shift staffing vs average staffing for the entire stay). The differing timeframes for the staffing data may have influenced the results.


Most adverse events (58%) were of the more severe ‘rescue’ category. Four-fifths of the adverse events occurred on off shifts (evening, night or weekend), while 54% occurred in the first postoperative day. Nurses assessed the vital signs or sedation depth of postoperative paediatric patients three to four times every 8 h. The adverse event group had lower nurse staffing (3 vs 3.4 h/shift) and twice the prevalence of respiratory, cardiovascular, neurological and musculoskeletal comorbidities. A positive association between surveillance and adverse events suggests that nurses increased surveillance in recognition of patients’ clinical deterioration. Staffing moderated (ie, influenced the strength of) the relationship between patient factors and surveillance; when staffing was low, patient comorbidity influenced surveillance. Surveillance moderated the relationship between staffing and adverse events but did not mediate (ie, account for) it.


This study was innovative in focus and method. It revealed that the relationships between nurse staffing, nursing surveillance and patient outcomes are more complex than theorised. Surprisingly, surveillance did not account for the association between staffing and outcomes. Instead it had a moderate influence; at low surveillance levels, staffing had no association with adverse events. Meanwhile, staffing explained only 15% of the variance in surveillance, suggesting other factors that influence surveillance deserve study.4 The finding of a positive association between surveillance and adverse events, while clinically reasonable, appears contrary to the notion that surveillance prevents adverse events.

A related advance in this field is the description of missed nursing care and its role in the nursing system link to patient outcomes. Theoretically, higher staffing increases the opportunity for surveillance and completion of all necessary care. Evidence about surveillance or missed care illuminates the quality of nursing care, which has proven difficult to measure. Linking the nursing care process to patient outcomes is significant because it will guide management efforts to improve nursing care quality.

The study's case–control design is efficient for achieving experimental control and testing complex relationships. Another innovation was the use of the electronic record to measure a core nursing function—surveillance. Single-institution studies such as this one have the advantage of rich data that permit a more granular exploration than multihospital studies, although the findings have limited generalisability. A final advantage of this study is that it provides evidence linking nurse staffing to adverse events in paediatrics, an understudied population.


View Abstract


  • Competing interests None.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.