Background Aged residential care (ARC) facilities throughout New Zealand will experience a higher demand over the subsequent decades, with many older people presenting with multimorbidity. In 2015, the International Resident Assessment Instrument for Long-Term Care Facilities (interRAI-LTCF) became the primary assessment tool nationwide in ARC facilities, making it possible to monitor and benchmark health outcome profiles between ARC facilities. Falls, pressure injuries (PI), and unintentional weight loss are geriatric conditions that are largely preventable when risk factors are properly managed. Nurse Maude ARC hospital, premier provider in Canterbury and invested in continuous quality improvement, tasked this research with clear priorities to improve the monitoring and reporting of their interRAI data, particularly targeting falls, PI, and unintentional weight loss.

Aims The primary aim of this study was to benchmark Nurse Maude’s interRAI-LTCF data with other hospital-level ARC facilities nationwide and in the Canterbury District Health Board (DHB) region, targeting falls, unintentional weight loss and PI. In doing so, this research would establish a procedure that could be used by Nurse Maude to carry out benchmarking on a range of health outcomes and indicators, in relation to themselves (over time), and national and Canterbury indicators. The secondary aim in this research was to determine if there are salient advantages in using the benchmarking procedure established in this research that offer Nurse Maude more than what is included in the two quarterly reports produced by Technical Advisory Centre (TAS). In the context of this research, benchmarking refers to a snapshot comparison at a point in time against two datasets that are expected to be similar (e.g., hospital-level of care).

Methods A cross-sectional study, of older people (aged 65 years and older) residing in hospital-level ARC that completed an interRAI-LTCF assessment between 1 July 2016 and 1 July 2017, inclusive, and who had consented to their data being used for planning and research purposes. Prevalence and risk of falls, unintentional weight loss, and PI were elicited within the interRAI-LTCF assessment, using the relevant assessment items, Clinical Assessment Protocols (CAPs) and Pressure Ulcer Risk (PUR) scale. Descriptive statistics were performed, and two-way cross-tabulation was conducted using Fisher’s exact test to determine that there was no statistically significant difference between the frequencies of interRAI-LTCF items/variables compared between datasets. A binominal test of significance was conducted to assess the probability of obtaining spurious significant result when comparing each dataset. The quality indicators for prevalence of falls, unintentional weight loss, and PI were calculated utilising the TAS definitions (i.e., numerator, denominator and exclusions). Nurse Maude interRAI-LTCF data was extracted through the interRAI analytics software, and the comparative Canterbury and national interRAI-LTCF data were provided by TAS. Clearance for this study was approved by University of Canterbury Human Ethics Committee. This study only required a low risk application given the use of de-identified routinely collected administrative data, which poses very low risk to the participants, community and research team.

Findings This study includes interRAI-LTCF assessments from 35 Nurse Maude hospital-level residents, as well as 1,698 and 14,021 residents from comparative Canterbury and national hospital-level datasets, respectively. This study found that Nurse Maude had a similar prevalence of falls (20%) and unintentional weight loss (11%) as national and Canterbury hospital-level datasets. However, Nurse Maude residents had a lower prevalence of PI (3%), than national and Canterbury hospital-level datasets (5% and 6%, respectively), and a higher risk of undernutrition (52% compared with 40% and 44% in Canterbury and national hospital-level datasets, respectively). In general, there was little variation between national and Canterbury hospital-level datasets for most interRAI-LTCF outcomes measured in this study.

Discussion Nurse Maude appeared to perform the same as or better than the national and Canterbury datasets (e.g., PI and falls quality indicators), but had slightly lower performance in other areas (e.g., undernutrition risk and low BMI score). However, the relatively small sample, which results in increased variability, is likely to be responsible for some of these patterns. Repeated analysis over time, and then employing using data stabilisation methods, would mitigate this issue and is recommended for the future. Overall, the benchmarking procedure established in this research provides some additional advantages for Nurse Maude compared with the TAS reports. These advantages include the real-time data access, indicator flexibility, supplemented missing weight and height data, and potential to use data stabilisation over time.

Implications These conditions remain prevalent in hospital-level ARC facilities throughout New Zealand, reinforcing the need for targeted, evidence-based initiatives to reduce the risk and prevalence of these modifiable and costly health care outcomes. This research established a procedure that can be used by Nurse Maude to carry out benchmarking on a range of health outcomes and indicators, in relation to themselves (over time), and national and Canterbury datasets. This process could complement the information provided to Nurse Maude in the TAS quarterly reports.