Article Text
PDF
Statistics from Altmetric.com
- Accidental falls
- hospital care populations/contexts
- economic evaluation
- cost-effectiveness
- randomised trial methodology
- older people populations/contexts
- clinical care interventions
- fall
- clinical care
- mixed methods
- hospital care
- public health
Background
Economic evaluations play an important role in decision making and the allocation of resources within the healthcare sector. Resources in hospitals are scarce, consequently choices are often made between competing alternatives. The resource implications of developing and implementing new programmes are often high, and justifying the costs required for their implementation is an ongoing challenge for clinicians and managers. It is therefore essential to incorporate the analysis of costs and benefits into the evaluation of prevention programmes. An economic evaluation aims to assist in the decision making process and ensure that programmes implemented have benefits that outweigh their opportunity costs.1 They assess the cost efficiency of a programme and investigate whether a programme makes the best use of the limited resources that are available.2 The most common and widely used approach to economic evaluation in healthcare is cost-effectiveness analysis.3 This paper describes the methods of an economic evaluation to be conducted alongside a cluster randomised controlled trial (RCT) to determine the cost-effectiveness of a falls prevention programme—the 6-PACK—from an acute hospital perspective.
In-hospital falls are the most common hospital adverse event in the acute hospital setting4 (38% of all reported incidents).5 Up to 30% of all hospital falls have been reported to result in an injury.6 They are associated with a substantial increase in morbidity and mortality, increased risk of institutionalisation and increased nurse workloads.7 Commonly identified risk factors for in-hospital falls include age, female gender, altered mental state, a history of falls, recent changes to medications, limited or impaired mobility and special toileting needs.6 8
Falls result in additional hospital costs, through their impact on length of stay and resource utilisation, such as medical investigations, care complexity and decreased patient independence.9 10 Studies have indicated that an in-hospital fall may result in an increase in patient length of stay up to 12 days10 and a doubling of hospitalisation costs9 compared with a non-faller. Therefore, in-hospital falls impose considerable economic consequences and are a potentially preventable source of hospitalisation costs. However, despite a considerable increase in the use of falls prevention programmes over the last decade, the evidence to support sustainable and cost-effective approaches in an acute care setting is sparse.
A variety of falls prevention programmes have been implemented within hospitals, with the aim of reducing the incidence of patient falls. However, a recent Cochrane systematic review and meta-analysis found limited evidence to support the effectiveness of such programmes in the acute hospital ward setting.11 A cost-effective falls prevention programme will have major economic benefits for hospitals and healthcare organisations, by avoiding the costs incurred by fall-related care,12 or benefits in terms of the reduction in the number of falls.
To our knowledge, there are no published economic evaluations that have been conducted alongside controlled trials of falls prevention programmes in the hospital setting.1 11 Two studies, however, have attempted to quantify the costs of implementation of a falls prevention programme using an economic modelling approach.13 14 Boswell et al conducted a retrospective epidemiology study to evaluate the impact of a patient sitter programme for reducing in-hospital falls.13 However, this study found no reduction in falls over the observation period and hence no economic benefit. Haines et al modelled the economic consequences of providing a patient education programme to all patients, to no patients, or to a subset of patients identified by their treating physiotherapist as being at risk of falls.14 The authors reported a reduction of 2.2 fallers and a cost saving of $A2704 per 100 patients if the education programme was provided only to the subset of patients identified by their treating physiotherapist as being at risk of falls. An assumption of this modelling was that education was equally effective across all subgroups of patients (based on results from an earlier RCT),15 16 but this assumption was shown in a more recent RCT to not be the case.17 While these studies provide some important insights into the costs of in-hospital falls prevention programmes, it is important to appreciate the limitations of modelling in economic evaluations. Both studies used cost estimates modelled from prior studies conducted up to two decades previously. In addition, there is a limited evidence base to support the effectiveness of the interventions studied.
Economic evaluations conducted alongside an RCT, comparing interventions incorporated into usual care practice, are accepted as the gold standard approach.3 However, the conduct of economic evaluations alongside falls prevention clinical trials in the hospital setting has to date been overlooked.1 This study aims to address this important evidence gap and is planned to be the first economic evaluation alongside an RCT involving a targeted falls prevention intervention to reduce in-hospital fall-related injuries—the 6-PACK programme.
The 6-PACK programme
A 9-year observational study conducted at The Northern Hospital, Victoria, Australia, using a targeted nurse delivered programme designed specifically for acute wards—the 6-PACK (figure 1)—found a sustained reduction in fall-related injuries and that the programme was easily integrated into usual care.19 This 9-year study found a 25% decrease in fall-related injuries in the first year, and a 50% reduction in the second, with reductions sustained for 5 years. Despite the preliminary evidence of the positive impact of the 6-PACK programme, little is known about the cost-effectiveness.
The 6-PACK programme.
The 6-PACK cluster RCT is being conducted to establish the efficacy of the 6-PACK falls prevention programme.20 This economic evaluation is designed to investigate the cost-effectiveness of the 6-PACK programme compared with usual care practice. It is being conducted to inform decision making, by hospital administrators and clinicians, relating to resource allocation, practice change and 6-PACK programme investment; thus the perspective of the acute hospital will be taken. The primary outcomes of this study used in the cost-effectiveness analysis will be ‘cost or saving per fall prevented’ and ‘cost or saving per fall-related injury prevented’.
Methods
Design
This economic evaluation will be conducted as part of a multicentre single blinded cluster RCT—the 6-PACK project.20 This evaluation has been informed published guidelines for economic evaluations for falls prevention interventions.1 It will be conducted in 24 wards from six hospitals, across Australia. The economic evaluation will take the acute hospital perspective and aims to compare the 6-PACK fall prevention programme with usual care practice (current ward based falls prevention). The intervention and control groups will be compared to determine the relative cost-effectiveness according to two clinical outcomes: (1) falls prevented; and (2) fall-related injuries prevented.
A 1-year cost-effectiveness analysis will be conducted to correspond with the 12-month cluster RCT period, which has been described in detail elsewhere.20
Study population
All participating wards recruited as part of the cluster RCT will be included in the economic evaluation. All patients admitted onto the participating acute wards will make up the participants of the study. Eligible hospital wards include acute medical or surgical wards with an average patient length of stay of <10 days and where fall-related injuries have been identified as a problem.
Pairs of wards from each hospital will be matched by ward type and fall-related injury rates from 6 months preceding the study implementation, and then randomly allocated to the 6-PACK intervention (12 wards) or usual care control group (12 wards) using a web-based allocation programme (figure 2). As the unit of randomisation will be based on clusters (wards) and not individual patients, the unit of analysis for both outcomes and costs will be at the ward level.
6-PACK cluster randomised controlled trial and economic evaluation.
Intervention: the 6-PACK programme
The 6-PACK programme is a targeted nurse delivered programme designed specifically for acute hospital wards.20 The programme includes a nine-item fall risk assessment (TNH-STRATIFY) and the delivery of one or more of six nursing interventions (figure 1) to patients classified as high risk. The TNH-STRATIFY, a valid and reliable tool for identifying patients at risk of falling, will be conducted on admission and during each nursing shift throughout a patient's admission.18 The 6-PACK interventions will be applied to individual high risk patients at the nurse's discretion.
Each 6-PACK intervention ward will be required to provide access to falls prevention equipment: falls alert signs, low-low beds (a minimum of one low-low bed to three standard beds on medical wards, and one low-low bed to 10 standard beds on surgical wards), and bed/chair alarms (a minimum of three on medical wards and one on surgical wards). In addition, each participating hospital will be required to appoint a part-time falls prevention clinical leader (site clinical leader), who will be responsible for implementing and monitoring of the programme on the intervention wards. Site clinical leader activities will include training sessions for intervention ward staff, 15-minute ‘ward walk-rounds’, case reviews, compliance audits and feedback sessions.
Comparator
The control wards will continue with standard usual care practice relating to falls prevention. This study will not restrict wards from using previously performed or new falls prevention activities. Falls prevention interventions used by control wards will be monitored throughout the trial period through falls prevention practice audits and structured observation. Any identified changes in practice that occur during the intervention period, such as capital purchases, will be appropriately costed.
Sample size
The study is powered to detect a 30% reduction in fall-related injury rates in the year post-programme implementation, based on 5% statistical significance and 80% power.20
We anticipate a total of 510 fall-related injuries, from approximately 16 000 patients and 212 000 occupied bed days during the 12 month cluster RCT.19 21
Measures of clinical outcomes: identifying falls and fall-related injuries
Fall and fall-related injury rates are commonly used clinical outcomes in trials involving fall prevention interventions. Therefore, for the purpose of this economic evaluation the clinical outcome measures used will be ‘falls’ and ‘fall-related injury’ rates per 1000 occupied bed days.
A fall will be defined as an event resulting in a person coming to rest inadvertently on the ground, floor or other lower level.20 22 A fall event will be further classified as resulting in an injury or not. A fall-related injury will be defined as any physical damage resulting from a fall (including bruising, abrasions, lacerations and fractures).23 Making this differentiation between an injurious and non-injurious fall will be valuable in providing more specific data on the costs and benefits of the 6-PACK programme.1
Fall and fall-related injury data from participating wards will be prospectively collected as part of the cluster RCT using a multi-modal approach and have been described in detail elsewhere.20 Briefly, it will include triangulation of data from daily medical record audits, verbal reports from the nurse unit managers, and audits of hospital administrative and incident reporting databases.
Measures of costs: identifying the resource impact of the 6-PACK
Using the acute hospital perspective, this study sets out to quantify the costs of implementing the 6-PACK programme at the ward level to hospitals. The cost items that will be considered in the analysis are outlined in table 1. Total costs of implementing the fall prevention programme and all hospitalisation costs incurred by the 6-PACK participating wards will be included. Cost data will be gathered as part of the cluster RCT. The cluster approach to this RCT means that patient level data will be aggregated at the ward level to determine programme and hospitalisation costs. All costs will be expressed in Australian dollars.
Cost items for inclusion in the economic evaluation
Direct costs of delivering the falls prevention programme
6-PACK intervention costs will reflect those associated with the implementation, delivery and monitoring of the 6-PACK programme by the site clinical leader. Resource utilisation associated with the 6-PACK programme will be based on data obtained from multiple sources: routinely collected financial, administrative and training records, structured observation as part of the cluster RCT, and the site clinical leader resource utilisation diaries. In addition, each hospital will be required to complete a falls prevention practice audit at the beginning and the completion of the study to obtain information about existing falls prevention policies and programmes.
The site clinical leaders will be required to complete weekly resource utilisation diaries (available from the authors on request) to quantify 6-PACK programme implementation and monitoring activities. Diaries will detail resource use associated with time spent organising and conducting training, monitoring and feedback sessions and meetings related to the falls prevention programme. Staff costs will be based on standardised costs per hour and site clinical leader incomes and/or government award wages for a clinical nurse specialist or equivalent. Resource diaries and financial and administrative records will provide data on the cost of consumables.
Capital purchase costs, acquisition of falls prevention equipment, will be obtained through audits of hospital equipment and financial records, and the site clinical leader weekly resource diaries. In addition, daily medical record audits and structured observation of falls prevention equipment used, on control and intervention wards, will be conducted over a 12-month period as part of the cluster RCT. Purchase or rental prices and relevant depreciation rates will be used to calculate capital and equipment costs.
Hospitalisation costs
Total hospitalisation costs will include all costs incurred by the participating 6-PACK wards for the 12-month study period and any additional costs that occur during the patient's stay on participating study wards. Ward costs will include items such as nursing, impress pharmacy, consumables and overheads. Patient level costs will be aggregated at the ward level, and include allied health, medical care, pathology, radiology and theatre costs for the duration of time a patient is on a participating study ward. Hospitalisation costs of study participants will be obtained from both the ward general ledger and the hospital computerised clinical costing systems.24
It should be noted that to avoid double counting, nurse delivered falls prevention interventions delivered as part of usual care practice during daily patient care, on the control and intervention wards, will not be individually costed. These costs are incorporated within the hospital clinical costing data that will be used to calculate ward hospitalisation costs. As such, direct falls prevention programme costs will not include any ward based falls prevention programme activities, but will include only those attributable to the implementation and monitoring of the 6-PACK programme through additional staffing or resources, equipment costs and consumables.
Exclusion of trial costs
All research costs related to the clinical trial will be excluded from the costs data analysis, including costs associated with data collection, analysis and measurement of clinical outcome outcomes performed by trial staff.
Cost-effectiveness analysis
Economic evaluations aim to capture both the costs and benefits of a programme in one comparable unit, a cost-effectiveness ratio.3 Data will be analysed on an intention to treat (ITT) basis, with the clusters (wards) as the statistical unit. The outcome of the cost-effectiveness analysis will be ‘cost or saving per fall prevented’ and ‘cost or saving per fall-related injury prevented’ calculated from differences in mean total costs and effects in the intervention and control groups to generate an incremental cost-effectiveness ratio (ICER=∆C/∆E) (table 2).
Incremental cost-effectiveness analysis
A cost per fall prevented ratio will provide a crude ratio of cost effectiveness that can be easily comparable with other falls prevention programmes.1 However, prior evaluation of the 6-PACK falls prevention programme found a reduction in fall-related injuries, rather than falls.19 In addition, there are likely to be substantial differences in the costs associated with falls and fall-related injuries. As such, it will be beneficial to differentiate between falls and fall-related injuries and to determine the cost-effectiveness of the 6-PACK programme for both falls and fall-related injuries prevented.
The differences in mean effects will be calculated using the methods specified for the primary analysis of the cluster RCT using negative binomial regression analysis adjusted for clustering by ward and pairing of wards.20 Mean costs for both the intervention and control groups along with measures of variability and precision will be reported. Differences in mean costs will be calculated using a hierarchical model to adjust for clustering by ward and hospital.25
Primary analysis for costs and effects will be unadjusted; secondary analysis will be conducted adjusting for potential baseline predictive variables, such as age, cognitive impairment and admission source.
Sensitivity analysis
Both one-way and multi-way sensitivity analysis will be conducted to examine parameter variability and the robustness of the ICER; for example, using 95% CIs for the parameters of interest to the hospital, such as 6-PACK programme and hospitalisation costs and 6-PACK effect estimates.
Ethical and governance approval
Multicentre approval has been granted by Monash University Human Research Ethics Committee (project number: CF11/0229: 2011000072) as part of the cluster RCT. Site-specific ethics and research governance approval has also been obtained from local ethics committees at all participating hospitals.
Discussion
This protocol outlines the design of an economic evaluation that aims to investigate the cost-effectiveness of the 6-PACK programme from an acute hospital perspective. It provides a detailed statement of the planned analysis.
There are a number of advantages to conducting an economic evaluation alongside a clinical trial. Clinical trials can provide an efficient opportunity to collect economic data within a robust research setting, enhancing the internal validity of the economic evaluation.26 In addition, economic input throughout the trial planning, design, data collection and analysis allows the data to be collected in ways that meet the economic needs of the evaluation.3
The strengths of this planned economic evaluation include the multi-modal approach to identifying falls and fall-related injuries, the robust study design and the prospective collection of cost data. A multi-modal approach to identifying falls in the acute hospital has been identified as the most reliable approach to achieve complete case ascertainment of fall events.27 The 6-PACK trial incorporates features for controlling predictable sources of bias and confounding (ie, concealed random allocation to groups, quality control of data integrity, blinded outcome analysis, and ITT analysis).20 Prospective measurement of costs alongside the 6-PACK trial will include comprehensive data collection of site specific clinical costs data supplemented with weekly resource diaries, equipment registries and audits. This aims to provide more specific estimates of hospitalisation costs when compared with economic modelling techniques using nationally reported hospital costing data.
There are, however, some methodology challenges that should be noted when conducting an economic evaluation alongside a cluster clinical trial design. Cluster RCTs are becoming increasingly used in practice to assess interventions that are implemented at an organisational, clinical unit or geographical level.28 They are often conducted when individual randomisation is not practical or ethical. Despite the increasing use of this study design in practice, there remains challenging issues that have not been largely discussed in prior work in relation to conducting an economic evaluation alongside a cluster RCT.
The unit of analysis in an RCT is the unit of randomisation. Therefore in a cluster RCT, principles such as ITT and outcome analyses are applied at the level of the cluster. In this 6-PACK trial, the use of hospital wards as clusters provides a challenge in relation to the capture of hospitalisation costs and the ability to cope with potential imbalances in both cluster and individual patient characteristics. Analyses of costs and effects will therefore be adjusted for clustering by ward and hospital to manage these potential clustering effects and the likely heterogeneity of costs across wards and hospitals.25 28 Hospitalisation costs of participating wards will account for changes in resource utilisation and patient length of stay during a patient's ward stay. However, changes in total hospital length of stay or costs incurred outside of the patient's ward length of stay will not be included in the analysis. As such, hospitalisation costs collected as part of this evaluation may not capture all downstream costs incurred by a fall event. Hence, ward and hospital discharge destinations will be recorded and taken into consideration when reviewing the outcomes of this evaluation. Analysis at the ward level is therefore likely to provide a conservative estimate of the incremental costs or savings per fall or fall injury prevented. Heterogeneity in practice patterns, resource availability, unit prices and clinical costing across clusters will be explored and managed within the sensitivity analysis.
The 6-PACK study is being conducted across six hospitals within Australia. This multicentre study aims to provide results that are generalisabile and transferable across Australia. The target audience for this economic evaluation is hospital administrators and clinicians. As such this evaluation takes the perspective of the acute hospital. The acute hospital perspective will provide sufficient information to inform decision making surrounding the implementation and delivery of cost-effective falls prevention programmes. The results may also be of relevance to public hospital funding bodies and health insurers.
Conclusion
This is a protocol for an economic evaluation aimed at providing data on the cost effectiveness of the 6-PACK fall prevention programme. Information gained from this evaluation will represent the first prospective economic evaluation of the 6-PACK falls prevention programme conducted in the acute healthcare setting. It represents an opportunity to explore the potential economic costs and benefits of the 6-PACK programme. If the programme is found to be cost-effective, findings could potentially have a significant impact on the implementation of falls prevention programmes in Australia and internationally. Evidence of cost-effectiveness can be used to support the implementation of the 6-PACK programme as an efficient approach to falls or fall-related injury prevention in the acute care.
Acknowledgments
Special thanks are extended to Gigi Chan, Decision Support, Eastern Health for providing valuable information on the hospital clinical costing systems.
References
Footnotes
Funding National Health & Medical Research Council, Australia (APP1007627).
Competing interests None.
Patient consent This study is a cluster randomised controlled trial. As such hospital wards and not individual patients have been recruited to this study.
Ethics approval Multicentre approval has been granted by Monash University Human Research Ethics Committee (project number: CF11/0229: 2011000072). Site-specific ethics and research governance approval has also been obtained from local ethics committees at all participating hospitals.
Provenance and peer review Not commissioned; internally peer reviewed.