Can Nursing Information Systems Predict Falls in Hospitalized Patients?

Featured Commentary

Loretta Secco, RN, BScN, MN, PhD,
Evelyn Kennedy, RN, BScN, MEd,
Sheila Profit, RN, BScN, MAdEd

Purpose and Procedure

The purpose of this segment is to showcase an interesting, informatics-focused article from the literature that warrants attention by nurses trying to learn and use nursing informatics in clinical practice. The article selected is one that was valued as relevant and informative for the nursing readership. The structured abstract was prepared by M. Loretta Secco and the commentary was prepared by nurse experts, Evelyn Kennedy and Sheila Profit, members of the Nursing Information Research Group, Joint St Francis Xavier/Cape Breton University Nursing Program in Nova Scotia.

Article Title: Can nursing information systems data predict falls in hospitalized patients?

Question: For hospitalized acute care surgical and medical elderly patients can nursing information systems data effectively predict falls?

Article Used for Evidence-Based Practice:

Giles, L.C., Whitehead, C.H., Jeffers, L., McErlean, B., Thompson, D. & Crotty, M. (2006). Falls in hospitalized patients: Can nursing information systems data predict falls? CIN: Computers Informatics Nursing; 24(3):167–172.

METHODS

Design:

Retrospective analysis was performed using extracted data from hospital nursing information system data bases. Risk of fall was defined with a refined set of Units of Care (UOCs) identified using data from the previous year (2001). The total of 30,723 UOCs was refined to a list of 100 based on frequency of assignment by the nurse. The list was further refined by an expert nursing panel to 28 UOCs most predictive of a patient fall. Univariate logistic regression analysis was employed to evaluate the extent that each UOC predicted faller status (p value of 0.2) and significant Inspection of Care (IOC) formed a set of risk factor variables for further multiple regression analysis.

Setting:

The hospital was a 250-bed acute care hospital that provided acute medical and surgical care for people living in Adelaide, South Australia and special service for war veterans and war widows.

Participants:

The data included in the study were captured from 7167 patients admitted in 2002. Main outcome measures: The outcome of the study was fall versus no fall.

Nursing Information System:

Three hospital databases used in this study: They included the (1)“Homer” patient management system, (2) the Advanced Incident Monitoring System, and (3) the Nursing Information System, “Excelcare”. The Advanced Incident Monitoring System records information on incidents, including falls. Nurses use the Excelcare system to record needed nursing care per shift in UOCs. The Homer system captures patient demographic and admission details.

MAIN RESULTS

Significant multiple regression nursing units of care (NUC) (i.e., odds ratios > 1) associated with patient falls included patient safety, confusion, continence, medication, mobility, and sleep. Total number of UOCs identified also predicted patient falls with significantly greater risk of falls associated with more UOCs. Several of the UOCs most predictive included urinary incontinence, risk management for falls, impulsive behaviour, and safety level rating. The NUC that was most predictive of fall status was urinary incontinence management, followed by risk management for falls.

Conclusions

Specific UOCs and total number of UOCs extracted from nursing information systems effectively predicted falls among hospitalized acute care elderly patients.

COMMENTARY

The study by Giles et al (2006) is important because patient falls not only cause significant morbidity and mortality but are a source of extreme distress for patients, families, nurses and other health care providers. Because families often see patient falls as preventable, occurrence often results in loss of trust that negatively impacts on the therapeutic relationship nurses have worked to develop. The Registered Nurses Association of Ontario developed a best practice guideline focusing on prevention of falls in health care settings (2005).

Fall prevention programs begin with risk assessment or screening tools and, as stated by the author, several such tools are available to effectively screen for risk and are part of most Quality Management protocols. Nurses in acute care settings critically need interventions and tools to improve monitoring and quality of patient care without an associated increase in workload burden. Our aging population, retirements, and continuing nursing shortage indicate that an easing of the heavy workload is not likely in the foreseeable future. Giles et al (2006) refer to studies that suggest poor compliance with screening tool usage, perhaps due to time constraints. The significance of the Giles et al study is the potential for a computer-generated screening tool to highlight fall risk and improve patient outcomes, as well as a high likelihood of nurse compliance due to reduced workload. This is a rare but winning combination.

Giles et al (2006) also highlight another important point on risk of falls, i.e., that risk differs significantly in hospital compared with community settings. The authors accurately point out that study findings are not necessarily applicable to patients in other settings and that compatibility with other patient care data systems must be considered. Nurse computer literacy level, technical support, and resources are necessary to implement successfully. However the tools and findings of this study represent an innovative step toward improving quality of care through a nursing system intervention.

References

Giles, L.C., Whitehead, C.H, Jeffers, L., McErlean, B., Thompson, D & Crotty, M. (2006). Falls in hospitalized patients: Can nursing information systems data predict falls CIN: Computers Informatics Nursing; 24(3):167–172.

Registered Nurses Association of Ontario (2005). Prevention of falls and fall injuries in the older adult (Revised), Toronto, Canada: Author.

Title:

Commentary on Can Nursing Information Systems Predict Falls in Hospitalized Patients?

Authors: Loretta Secco, RN, BScN, MN, PhD, Evelyn Kennedy, RN, BScN, MEd, Sheila Profit, RN, BScN, MAdEd

E-mail: Evelyn_Kennedy@capebretonu.ca

Affiliation: Joint St Francis Xavier/Cape Breton University Nursing Program, Nova Scotia.

Submitted: June 2006
Accepted: August 2006

Editor: June Kaminski, RN MSN PhD St.

APA Reference:

Secco, L., Kennedy, E. & Profit, S. (2006). Commentary on Can Nursing Information Systems predict falls in hospitalized patients? Canadian Journal of Nursing Informatics, 1(2). https://cjni.net/journal/?p=331