The search strategy revealed 657 references from the following sources: PubMed (n = 345), EconLit (n = 14), and Google Scholar (n = 298). Fifty-one duplicate references were removed. Of the remaining 606 references, 581 were removed because they did not focus on QI (n = 392), were not conducted in LTC settings (n = 141) and not economic studies (n = 48). Based on a full-text screening, nine references were removed (conference abstracts, n = 4 and study protocols, n = 5). Sixteen references were included in this review, of which 13 were economic evaluation studies and three were systematic reviews (Figure 1).
Summary of the included studies
There were 13 individual economic studies and three systematic reviews. These studies originated from Canada (n = 3), United States (n = 2), United Kingdom (n = 2), Australia (n = 2) and the Netherlands (n = 2), along with one study each from Spain and Germany. The study topics represented a wide range of QI strategies such as pharmacist-led medication reviews [16-19], peer coaching for safe resident handling [20-21], preventive practices for reducing falls [22-23], preventing pressure ulcers [24-26], infection prevention [27-28], and dealing with challenging behaviors [29-30]. Several studies involved a multidisciplinary team, comprised of geriatricians, pharmacists, psychologists, registered nurses, licensed practical nurses, recreational therapists and care-aides working in the LTC settings. Most of these studies were published in the last five years (n = 11). Only 2 studies were published in the health economics-related journals, whereas other studies were published either in geriatric medicine or clinical medicine journals. Table 1 summarizes the included studies.
Methodological approaches
Table 2 lists methodological approaches in the included studies. The economic studies represented a wide range of methodological approaches such as trial-based economic evaluations (n = 5), cost-effectiveness studies (CEA) (n = 5), cost-benefit studies (n = 2) and a cost-minimization study (n = 1). A vast majority of the studies included analyzed primary data (n = 9), while a few studies reported on a retrospective analysis of the secondary data (n = 4) and systematic reviews (n = 3). Almost all studies examined financial costs relative to the health benefits of QI strategies using a public-payer (i.e., health care system) perspective. Because many studies evaluated QI projects over a short-time horizon, incremental costs and outcomes were reported for < 2 years. Only two studies applied a Markov-chain model to calculate long-term, 20 years or lifetime, costs and outcomes of QI strategies [24-25]. Studies originating from Canada used a discounting rate of 3 to 5% [21,24,26], as compared to 7% in other studies originating from the United States [20], and Germany [23]. The uncertainties in the model input parameters of costs and outcome were addressed by one-way deterministic sensitivity analysis [18,20-22,26,28], nonparametric bootstrapping [17], and probabilistic sensitivity analysis [16,23-25]. Only two studies reported on a budget impact analysis [23,26].
Financial costs of QI interventions
The cost inputs were derived from multiple sources including health insurance providers, national/provincial administrative databases, LTC facility records and costing exercises embedded alongside the trial. The financial costs were categorized into three broad-buckets: (i) developmental, (ii) project/program implementation, and (iii) health care. The developmental costs, often one-time expense, represented planning activities such as, initial consultation [18], focus groups [21], and training [18,29]. The project/program implementation included costs of equipment, maintenance [18,20,24,28,30], educational package [23,29], printing and supplies [16,21,26]. The health care costs comprised of administrative staff and care-providers’ salaries in the LTC facilities [18,23-24,26], transfers to hospitals [22], in-patient hospitalization [16,18,22,23], out-patient-and-ambulatory visits including day-surgeries/procedures [16,18,22-23,25], medical supplies and medications [16,23,28]. Only two studies reported on the societal costs (i.e., costs to the health system, as well as out-of-pocket costs and productivity losses) [21,29]. (Table 3)
Health outcomes/benefits of QI interventions
The literature revealed a combination of the economic and clinical outcome variables relevant to QI intervention. The terms ‘Quality of life’ (QoL), ‘Health-related Quality of Life’ (HRQoL) and ‘Quality of Care’ (QoC) were used interchangeably to express health gains. Studies conducted by Jódar‐Sánchez [17], and Twigg et al. [18] administered a validated tool (i.e. EuroQol EQ-5D-5L) to measure HRQoL – mainly quality-adjusted life years (QALYs) – in elderly residents undergoing pharmacist-led medication reviews. Other studies extracted disease-specific QALY estimates from previously published literature [23-25]. The QoC variables included infection prevention [27], falls or injury prevention [16,21-23], reducing the length of hospital stay [24], and avoiding hospital readmissions [22]. Some studies also reported other outcomes specific to a medical condition or disease such as cognitive improvements [18], reductions in the disease incidence or prevalence [25], preventing mortality [23-24], and reducing agitation-level [29-30]. Overall, the economic outcomes were presented as the incremental cost relative to benefit [16-18,22-25,29-30], net monetary benefit [20-21], and cost savings [19,26,28] or cost avoidance [4,26-27].
Cost-effectiveness of QI interventions
Zwijsen et al., [29] performed a trial-based CEA of GRIP- a challenging behaviour care program based on the recent guideline-driven educational package for managing elderly residents diagnosed with dementia, as compared with standard care in the Netherlands setting. The findings showed that implementation of the GRIP program was more costly than usual care (i.e., mean cost difference € 276; and 95% CI: €237 to €349), and there was a very small difference in the QALYs (i.e., -0.02; 95% CI: -0.06 to -0.003) between two groups. Another economic study from Spain evaluated the cost-effectiveness of a pharmacotherapy follow-up using a prospective observational research design [17]. In the base-case analysis, the pharmacotherapy follow-up was found more costly (crude unadjusted incremental cost, €399) and less effective (incremental QALYs, -0.001) as compared with usual care. However, the incremental cost-effectiveness ratios (ICERs) appeared more favorable (ranged from €3,898.69 to €6,573.56 per QALY gain, and the probability of pharmacotherapy follow-up being cost-effective was 76-78% at a willingness-to-pay (WTP) threshold of €30,000 per QALY) when both costs and utility scores were adjusted in the subsequent scenario-based analysis. A cost-benefit analysis conducted by Lahiri et al., [20] reported cost savings of implementing a safe resident handling program in the United States (US). This study also showed mixed results suggesting a substantial variability in net savings across LTC facilities. Of 110 facilities, 61 (55.5%) had a positive net savings of approximately US$3,064 per bed, and negative savings (i.e., the incremental cost of US$992 per bed) in the remaining 49 facilities. Stern A et al., [24] also showed an incremental cost of CA$731 and incremental QALYs of 0.03 (ICER: CA$7,824,747 per QALY gained) for the oral nutrition supplements in high-risk residents with recent weight loss. Additionally, the prevention strategy of skin emollient for high-risk residents with dry skin yielded an ICER of CA$78,286 per QALY gained. Since the ICERs were above the standard acceptable WTP threshold of CA$50,000 per QALY, both prevention strategies were not considered cost-effective.
Other studies demonstrated favorable ICERs (i.e., lower incremental cost relative to benefit). In a trial-based economic evaluation of the pharmacist-led reviews of psychoactive medications (specifically antipsychotics, hypnotics, and anxiolytics) as compared with usual care in the United Kingdom [16], the proportion of nursing home residents receiving inappropriate psychoactive drugs were 19.5% and 50.4% in the intervention and control groups, respectively. The average cost of health resource utilization was relatively lower ($4,923 per resident per year, 95% CI: $4,206 – $5,640) in the intervention group, as compared to $5,053 per resident, per year (95% CI: 4,328 - $5,779) in the control group. In this study, the ICER was estimated at – 422 (i.e., incremental cost per proportion of resident receiving one or more inappropriate psychoactive drugs). Hewitt et al., [22] found an ICER of AU$18, per fall avoided; 95% CI: - $380.34 to $417.85 for a strength and balance exercise program (SUNBEAM), as compared to usual care in a randomized control trial in Australia. In a study conducted by Muller et al. [23], the multifactorial fracture prevention program implemented by a multidisciplinary team was also found highly cost-effective, as compared to no prevention. This study showed the incremental cost of €119 and incremental QALYs of 0.006 (ICER: €21,353 per QALY gained), yielding a probability of cost-effectiveness of 90% at a WTP of €36,000 per QALY).