This study demonstrated that medication reviews performed by clinical pharmacists during ward rounds could significantly reduce hospital costs by preventing ADEs. Our results showed a cost-saving ratio or ROI of €1.71 for each €1 invested in clinical pharmacy activities. Our calculated benefits fall within the range of findings from prior studies.
Few studies evaluated the ROI of clinical pharmacist activities in a hospital setting, and even fewer quantified the avoided costs associated with ADE prevention or calculating a median avoided cost for each DRP category. To the best of our knowledge, this study is the first to use real institutional costs from analytical accounting through a micro-costing approach.
Major pharmaco-economic reviews on clinical pharmacy interventions showed a wide range of ROIs from 1 to 75.84 [15–18] reflecting the variability in study types included, interventions evaluated and costs considered. For example, research perspective adopted can impacts the costs considered.
In Schumock and Perez’s review the median ROI are ranging from 1 to 4.81 [16, 17], corroborating our own. Studies that monetized ADEs using costs per DRP category calculated an ROI ranging from 1 to 14 [8, 9, 21, 22, 45, 46]. All these studies used a slightly different DRP cost calculation method (e.g., fixed cost for any ADE; fixed ADE occurrence probability; costs saved directly when a drug was stopped or changed.)
Two European studies [22, 46] used the micro-costing method developed by Rottenkobler et al. [3] to assign an avoided cost to a prevented ADE associated with a probability of occurrence score, as per Nesbit et al. [21]. The ROIs measured were 1 to 8.64 and 1 to 1.76. They fixed a value of a prevented ADE to be € 1,057 and € 1,079 respectively. Rottenkobler’s cost calculation method was quite similar to ours in that it incorporated the sum of the single cost components associated with inpatient treatments for ADEs based on cost centres. This way of quantifying ADEs enables more accurate and transparent cost assessments.
The differences between prior studies illustrated the variabilities in healthcare systems, settings, types of costs, costs calculations, or definitions of a pharmaceutical intervention or an ADE. This makes data comparisons, data extrapolations and generalizations challenging.
Our study's strength lies in method of avoided cost calculation, which was closer to the reality on wards than what other researchers have used up to now. We put values on ADEs using real costs derived from our hospital’s in-house accounting data rather than estimating costs from the literature. This was innovative because we calculated a median avoided cost for five DRPs category by costing real clinical situations at risk of ADEs and prolonged hospital. The actual hospital cost of managing each of these situations could be calculated by translating ADE into a disease or condition whose cost was easily quantifiable.
Costs were saved in each category of DRP detected by the pharmacists, except drug use without an indication. Potential ADEs from this DRP category were calculated to have a median cost € 0 because in more than 50% of the 20 cases quantified, the potential ADEs would only occur in the long term (e.g., osteoporosis, clostridium difficile or respiratory infections caused by proton pump inhibitors) without additional costs during hospitalisation. These pharmaceutical interventions remain relevant and could be valuable to studies examining ROI calculations from a societal perspective. It is worth noting that DRPs detection also aims to impact clinical and human factors. Although identifying a DRP may not only save money, it can improve the patient’s quality of life.
Sensitivity analyses showed that our economic model was robust. The break-even point demonstrated that intervening solely on low-probability or low-cost ADEs was worth the effort financially.
Our study suggests that pharmacists need to prioritize their interventions on risky DRPs to avoid drug-related morbidity costs. They should focus their interventions primarily on patients who pose a high clinical and financial risk to the hospital, regardless of whether they are common or uncommon. (e.g., fragile populations, poly-morbidity/medication).
Pharmacists should also focus on drugs with a high risk of acute ADEs or medication errors (e.g., drugs with narrow therapeutic range). To achieve this, they should seek the support of digital tools to help identify these situations. New existing systems can flag patients at risk of ADEs by using different triggers in electronic patient records. Moreover, they are cost-effective [47–51].
Unlike other studies [8, 9, 21, 31, 39, 52], ours did not include direct costs saved in medication (e.g. an inpatient’s prescribed expensive compound to a less expensive one, switching from an intravenous to an oral form, or using a biosimilar instead of an original drug). If other types of costs as humanistic, indirect, and intangible costs were considered in this analysis, the savings would have been even increased.
Despite our innovative cost calculation method, the present study had some limitations. The data sample for measuring costs associated with different categories of DRPs was limited. A bigger sample size would have resulted in more accurate, realistic costs. Random data selection from such a small sample may not be sufficiently representative of the actual distribution of identified risk situations. The expert panel may also have influenced the costs used since other evaluators might have chosen differently at this stage of the project. The precise attribution of avoided cost to a DRP is highly variable because it depends on the DRP’s actual clinical impact, which is difficult to predict. For example, we calculated the savings from detecting a drug interaction DRP at € 725. Other studies estimated these avoided costs at between € 285 and € 14,943 [31, 33, 34, 38]. The way in which costs are quantified can greatly influence this value. The precise definition of pharmacists’ clinical interventions and categories of DRP may also influence cost calculations and could explain inter-study cost variations. There are no standardized definitions of these interventions.
Conducting a comparative cost-benefit study is recommended to assess the best financial impact of clinical pharmacists. It will enable an estimation the actual differences in hospitalisation costs between inpatients with DRPs undergoing or not medication review. Previous research teams have already conducted cost-benefit analyses and cost-minimization studies to evaluate the cost-effectiveness of pharmacists' interventions [29, 47, 53–55]. Many cost-effectiveness publications have utilized cost-utility studies to measure benefits in quality-adjusted life years [35, 54, 56–63]. Alternatively, using prospective randomized controlled studies or before-and-after economic studies could enhance the measurement of avoided costs, as demonstrated by some research groups [52, 53, 62, 64–66]