The effectiveness of a multidisciplinary QI activity for accidental fall prevention: Staff compliance is critical

Accidental falls among inpatients are a substantial cause of hospital injury, resulting in extended lengths of stay and a decline in quality of life [1‐3], and occur in 3-20% of inpatients [4]. Starting on October 1, 2008, the United States’ Center for Medicare and Medicaid Services began including inpatient falls as a‘never event’—i.e., a preventable condition that should not occur after admission to the hospital [4]. Given that hospital falls are a universal occurrence, many of which are preventable, in-hospital falling remains a major healthcare concern internationally [5]. In recent randomized controlled trials, risk assessment tools and intervention plans have often been combined and studied as part of multifactorial intervention programs which include various modalities, including exercise, medication modification, vitamin D supplementation, environment/assistive technology, and both staff and patient education [6‐10]. Though data from randomized controlled trials and high-quality systematic reviews are considered the highest level of evidence, the evidence-practice gap remains substantial [11‐13]. Although a number of successful experimental studies on fall prevention have been published, the degree of effectiveness varies across studies[7‐9]. This is even more surprising given the many similarities among multidisciplinary fall prevention protocols. In an non-experimental clinical setting, Schwendimann et al., conducting an observational study to investigate the effect of an interdisciplinary falls prevention program, failed to show a substantial decrease in either frequency of falls or consequent injuries, citing low staff compliance as a probable cause [14]. While this suggests the importance of staff compliance with effective, but potentially cumbersome, multifactorial interventions, the role of staff compliance with prevention protocols has not been fully investigated in a routine clinical setting.

Quality improvement (QI) activity typically includes staff education, serial audit and feedback, and may even include financial incentives. It includes multiple modalities, but is often most successful when it employs a multidisciplinary, continuous, and systematic approach to hospital quality and care [15]. In particular, high rate of compliance with intervention plans may be achieved through serial audit and feedback sessions[16]. At our hospital prior to 2004, assessments and intervention plans varied from ward to ward and were subject to neither strict monitoring nor standardized reporting. The purpose of this observational study was to assess the efficacy of a multidisciplinary QI activity for accidental fall prevention, with particular focus on staff compliance in a routine, non-experimental, clinical setting.

During our 7-year study period, a total of 109816 patients were hospitalized. Among these, 71396 patients remained for risk assessment after exclusion of maternity, preventative health screening patients, intensive care patients, and patients who stayed in hospital less than 24 hours. Since the initiation of standardized risk assessment in 2007, 16829 patients (23.6%) were identified to be at risk for falls. Patient characteristics, event, and compliance data are summarized in Tables 4 and 5.

The overall fall rate was 2.13 falls per 1000 patient days (350/164331) in 2004 versus 1.53 falls per 1000 patient days (263/172325) in 2010, representing a significant decrease (p = 0.039). Bone fracture rates due to falls among hospitalized patients declined, though not significantly, from 0.04 fractures per 1000 patient days (6/164331) in 2004 to 0.02 fractures per 1000 patient days (4/172325) in 2010. Compliance with use of the falling risk assessment tool at admission was 91.5% in 2007, representing the latter half of the year with assessments beginning in July, and increased to 97.6% in 2010. The rate of implementation of an appropriate intervention plan was 89.6% in 2007, increasing to 95.3% in 2010. While this varied substantially by ward in 2007, all wards achieved close to 100% compliance (range: 98.5%-100%) in 2010.

This observational study assessed fall rate over time, consequent injuries, and characteristics of hospitalized patients in the periods before and after QI activity implementation. In addition, staff compliance with risk assessment on admission, as well as implementation of risk-stratified intervention plans, was examined. Before QI activities were implemented at our facility in 2007, the accidental fall rate had remained constant at approximately 2.00 falls per 1000 patient days. This study documents a 25% reduction of inpatient falls over five years, from 2.13 falls per 1000 patient days in 2004 to 1.53 falls per 1000 patient days in 2010, with the most dramatic reduction from 2006 to 2009. The consequent injury rate was less than 0.1% for any year. Feasibility of this integrated approach was excellent, as reflected by high staff compliance with use of assessment tools and implementation of intervention strategies across wards and departments.

Our initial fall rate at the beginning of the study period was lower than other studies in urban, acute-care hospitals, though in line with the lower end of the reported spectrum, typically between 2.2 falls per 1000 patient days to 6.3 falls per 1000 patient days [20‐22]. While the interventions put in place in 2006 were effective, aiding the further decrease was that the assessment rate of newly admitted patients was nearly 100% in 2010. Even if a robust preventive program exists, effectiveness is unlikely if compliance is low. A previous, observational study by Schwendimann et al. failed to show a substantial decrease in either frequency of falls or consequent injuries following the implementation of a similar interdisciplinary fall prevention program; the authors state that this was likely due to low staff compliance [14]. In contrast, a 2010 randomized controlled trial by Dykes et al. achieved a 46% reduction in their intervention group compared to a 25% reduction in the control group. Their staff compliance rate with intervention was 81%, corroborating the success in our study, which saw a final compliance rate of 95.3% [7]. This suggests that robust interventions themselves are necessary--but not sufficient--for fall prevention. Staff compliance with effective interventions is a critical, and perhaps often overlooked, factor in further closing the evidence-practice gap.

To this end, our QI activity included three elements to purposefully promote staff compliance with the fall prevention protocol. First, the Working Group in our hospital systematically involved healthcare staff on both planning and implementation levels in a multidisciplinary strategy to address the problem of accidental falls. The involvement of staff from multiple clinical and ancillary disciplines at the highest levels of planning may have influenced compliance at later points “downstream.” Second, all fall events were systematically documented, and subsequently followed by consistent short- and long-term auditing and feedback provided to hospital, department, ward, and individual staff at regular intervals. Finally, evidence-based reviews and practical training in protocol use were provided regularly to all staff in the form of repeated, evidence-based educational programming.

This study has some limitations. First, there was too little data on bone fractures after falls, thus we could not investigate the influence of our multidisciplinary QI activity on the rate of reduction of bone fractures. Second, this study was conducted in a single institution in a Japanese acute care hospital, with consequent uncertainty about generalizability. Future multi-center studies with larger numbers of patients may allow us to better clarify both effectiveness in reducing consequent injury, as well as assess the feasibility of our QI activity and result generalizability to other hospitals. The goal of our study was to decrease inpatients fall hospital-wide. Application of this protocol to those at highest risk, such as inpatients on geriatric or long-term nursing care wards, may be useful. Finally, it is difficult to distinguish the role that compliance plays in effective QI activity versus elements of the activity itself. Our initial compliance rate was already high and this may not be the case at other institutions. Our results, however, suggest that both are necessary components of QI protocols and should be maximized during planning phases. While we achieved a substantial increase in compliance rates after beginning QI activities, compliance was not well-assessed prior to its implementation. It is possible that a pre-existing “culture of compliance” contributed to the observed success. Effective ways to create and maintain this “culture of compliance” is an important future area of QI research and may vary by country, culture, and institution, and thus may not be as easily achieved in all practice settings.

Following the implementation of an integrated, multidisciplinary QI strategy for fall prevention in a 520-bed tertiary, acute-care hospital, we observed a 25% reduction in-patient falls with excellent staff compliance. This success, compared to variable outcomes in similar prior studies with lower compliance rates, strongly suggests that staff compliance with QI intervention plays a critical role in its efficacy. High staff compliance was facilitated by multidisciplinary staff involvement in QI activity planning and implementation; daily outcomes monitoring and monthly, ward-specific feedback regarding incidence of falls, consequent injuries, standardized risk assessment rate at admission, and rate of implementation of appropriate intervention; and practical education programs provided to all staff on a regular basis. To reduce the incidence of in-hospital falls, a systematized QI approach that closely involves, encourages, and educates healthcare staff at multiple levels is integral to its effectiveness.

This work was supported by the St. Luke’s Life Science Institute Research Fund [grant number: 2008-0-7].

Clinical Research Grant from St. Luke’s Life Science Institute.