Economic Evaluation of Pharmacist-Led Digital Health Interventions: A Systematic Review

There has been growing interest in integrating digital technologies in healthcare. The purpose of this study was to systematically review the economic value of pharmacist-led digital interventions. PubMed, Web of Science, and the Cochrane databases were searched to select studies that had conducted economic evaluations of digital interventions by pharmacists for the period from January 2001 to February 2022. Economic evidence from 14 selected studies was synthesized in our analysis. Pharmacists used telephones, computers, web-based interventions, videotapes, smartphones, and multiple technologies for their digital interventions. Prior studies have reported the results of telephone-based interventions to be cost-effective. Alternatively, these interventions were found to be cost-effective when reevaluated with recently cited willingness-to-pay thresholds. In addition, pharmacist-led interventions based on computers, web-based interventions, smartphones, and multiple technologies have been reported to be cost-effective in previous studies. However, videotape-based intervention was found cost-ineffective because there was no significant difference in outcomes between the intervention and the usual care groups. If this intervention had been intensive enough to improve outcomes in the intervention group, favorable cost-effectiveness results could have been obtained. The economic evidence in the previous studies represented short-term economic values. Economic evaluations of the long-term value of digital interventions are warranted in future studies.


Introduction
Digital health is a discipline that applies software and hardware to enhance health and health care delivery [1,2]. According to the FDA, the scope of digital health widely includes mobile health, telehealth (or telemedicine), wearable devices, and health information technology [3]. Recent advances in digital technology offer the opportunity to broaden the reach of healthcare. For example, increasing numbers of individuals in rural areas with limited access to healthcare can now receive timely interventions from remotely located healthcare providers. Multiple physicians can easily access patient charts, communicate with patients, and monitor disease management using their smartphone, mobile applications, and accessible software [4][5][6][7][8]. Pharmacists also use digital technology in their practices. They used telephones, mobile applications, and a remote monitoring device to improve medication adherence and drug-related outcomes, to educate patients, and to monitor patient symptoms remotely [9][10][11]. Pharmacists also provided computer-or web-based interventions to improve hypertension management [12] and to reduce medication errors and adverse events [13,14]. Sometimes, they used multiple technologies in their interventions [15,16]. Specifically, pharmacists offered face-to-face consultation (based on either a web-based program or a technology platform) plus telephone follow-up to communicate with patients, to educate patients, and to manage their health-related conditions in an effective manner. Because of the benefits conferred by technology, digital interventions have gained increasing popularity in healthcare.
By providing information of the incremental impact on economic and health outcomes as compared with alternative interventions (e.g., face-to-face usual care), economic evaluations of digital health interventions can guide practitioners or policy makers on whether to adopt such interventions. Digital interventions may be expected to bring economic benefits, for example, by reducing transportation costs, by decreasing time to diagnosis, and by making it possible for patients to remain in their homes rather than resorting to costly health care facilities [17,18]. In contrast, these interventions may prove costly because of the high upfront expenses for implementation, ongoing costs for support and maintenance, increased expense for added personnel interventions, and their resource-intensive nature [19,20]. Therefore, uncertainty exists about whether technology-based interventions are cost-effective. For this reason, studies have examined the economic value of digital interventions such as telehealth/telemedicine, electronic health (e-health), and mobile health (m-health) [21][22][23][24][25][26][27]. These studies have shown mixed results on the cost-effectiveness of digital interventions. Economic value of the interventions has varied, for example, by study populations, types of interventions, comparison groups, and outcome measures. Of note, all of these studies have focused primarily on physicians' interventions for specific population groups such as patients with diabetes [23,24], somatic conditions [25], or mental illness [26], and adolescents with chronic conditions [27]. To date, no studies have reviewed the cost-effectiveness of pharmacist-led digital interventions in general, regardless of disease type. Therefore, questions remain as to the potential economic impact that pharmacists would be able to exert on patients' outcomes. Accordingly, the objective of this systematic review was to synthesize the economic evidence of digital interventions provided by pharmacists for their patients. Our review provides economic evidence for multiple digital technologies that can help practitioners make decisions in adopting these technologies. In addition, such evidence can contribute to a better understanding of the benefits of extensive investment of these digital tools. Information from economic evaluation can be also used to make coverage and reimbursement decisions.

Search Strategy
A systematic literature search was conducted to identify all published articles assessing the cost-effectiveness of pharmacist-led digital interventions using databases such as PubMed, Web of Science, and the Cochrane database of Systematic Reviews for the period from January 2001 to February 2022. Here, digital health was operationally defined as the use of information and communication technologies to improve health, and thus, it included categories such as mobile health, telehealth, electronic health as well as wearable devices and any other tools that could be used to enhance health. To guide the search strategy, the following patient, intervention, comparison, outcome (PICO) framework was developed: P: Adult patients I: Digital intervention(s) by pharmacists C: Usual care (i.e., comparator treatment in published studies) O: Health-related clinical outcomes from patients The Medical Subject Heading (MeSH) terms and keywords for pharmacist-led digital interventions (pharmacist digital intervention, pharmacist mobile health intervention, pharmacist telehealth, pharmacist telecare, pharmacist teleconsult*, pharmacist telecommunication, telepharmacy, pharmacist electronic health, pharmacist ecommunication, pharmacist remote consultation, pharmacist short message service, pharmacist text messag*, pharmacist internet consultation, pharmacist internet monitoring, pharmacist communicate*, pharmacist video consultation, pharmacist vide monitoring, pharmacist video communicate*, pharmacist mobile application intervention) were coupled with those terms meaning economic evaluation (comparative effectiveness, economic evaluation, economic modeling, cost-benefit, cost-effectiveness, cost utility, cost-saving*, pharmacoeconomics). The search was limited to full-text articles published in English. Figure 1 depicts a flow diagram showing the selection process. The search strategy identified 2448, 38, and 7 articles from PubMed, the Cochrane database, and Web of Science, respectively. Removing duplicates left a total of 2491 unique articles. The titles and abstracts of these articles were reviewed in the first level of screening. Articles were excluded if they were (a) irrelevant to the purpose of this review; (b) pilot or proof-of-concept studies; (c) qualitative studies; and (d) not full text articles (e.g., poster, abstract, or letter). This screening process resulted in the exclusion of 2362 articles. The remaining 129 articles were then assessed for eligibility. Articles were excluded if (a) no digital interventions were provided to patients; (b) interventions were not delivered primarily by pharmacists; and (c) no cost-effectiveness metrics (e.g., incremental cost-effectiveness ratio) were reported. After applying these exclusion criteria, a total of 14 articles remained and included in our analysis.

Quality Assessment of the Studies
The Quality of Health Economics Studies (QHES) instrument, a validated tool to assess the quality of economic studies [28], was used to evaluate the quality of the 14 selected studies. As presented in Table 1, this instrument includes 16 items, each of which has a weighted score ranging from 1 to 9. The sum of the total item scores ranges from 0 (lowest quality) to 100 (highest quality). Using the QHES instrument, the quality of the studies included in this review was determined as poor (QHES score < 50), fair (50 ≤ QHES score < 75), or good (75 ≤ QHES score ≤ 100), as suggested by prior studies [29][30][31].
Two reviewers (H-M.K. and S.S.) independently extracted data from the 14 studies and evaluated their quality. The two reviewers showed a high level of agreement (ranging from 82.1% to 95.5%) in extracting data and evaluating the quality of the studies. Any discrepancies were discussed with a third reviewer (T.P.) to reach consensus.

Questions Points
Records identified through database search

Quality Assessment of the Studies
The Quality of Health Economics Studies (QHES) instrument, a validated tool to assess the quality of economic studies [28], was used to evaluate the quality of the 14 selected studies. As presented in Table 1, this instrument includes 16 items, each of which has a weighted score ranging from 1 to 9. The sum of the total item scores ranges from 0 (lowest quality) to 100 (highest quality). Using the QHES instrument, the quality of the studies included in this review was determined as poor (QHES score < 50), fair (50 ≤ QHES score < 75), or good (75 ≤ QHES score ≤ 100), as suggested by prior studies [29][30][31].

Questions Points
1. Was the study objective presented in a clear, specific, and measurable manner? 7 2. Was the perspective of the analysis (societal, third-party payer, etc.) stated and were reasons for its selection either stated or implied? a,b 4 3. Were variable estimates used in the analysis from the best available source (i.e., randomized control trial-best, expert opinion-worst)? c 8 4. If estimates came from a subgroup analysis, were the groups prespecified at the beginning of the study? 1 5. Was uncertainty handled by statistical analysis to address random events or sensitivity analysis to cover a range of assumption? 9 6. Was incremental analysis performed between alternatives for resources and costs? 6 7. Was the methodology for data abstraction (including the value of health states for cost-utility analysis and other benefits) stated? b 5 8. Did the analytic horizon allow time for all relevant and important outcomes? Were benefits and costs that went beyond 1 year discounted (3% to 5%) and justification given for the discount rate if the discount rate was arbitrarily determined? b 7 9. Was the measurement of costs appropriate and the methodology for the estimation of quantities, unit costs, and price years clearly described? a,b 8 10. Were the primary outcome measure(s) for the economic evaluation clearly stated and did they include the major short-term, long-term, and negative outcomes? a 6 11. Were the health outcomes measures/scales valid and reliable? If previously tested valid and reliable measures were not available, was justification given for the measures/scales used? 14. Did the author(s) explicitly discuss direction and magnitude of potential biases? a 6 15. Were the conclusions/recommendations of the study justified and based on the study results? 8 16. Was there a statement disclosing the source of funding for the study? 3 Total points 100 a Weighted partial points were awarded if a study yielded at least one "yes" to the questions in the item. b Authors' modifications are indicated by italics. c Weighted partial points were awarded if variable estimates were obtained from several sources with different qualities.
Two reviewers (H-M.K. and S.S.) independently extracted data from the 14 studies and evaluated their quality. The two reviewers showed a high level of agreement (ranging from 82.1% to 95.5%) in extracting data and evaluating the quality of the studies. Any discrepancies were discussed with a third reviewer (T.P.) to reach consensus. Table 2 shows an overview of the characteristics of the 14 studies included in our review. Of the 14 studies, cost-effectiveness analysis (n = 6) and cost-utility analysis (n = 6) were the most frequently used types of economic analyses. These studies were conducted in seven different countries including the U.S. (n = 5), Spain (n = 3), and Australia (n = 2). Various perspectives (patient, payer, and societal) were adopted in the studies. The majority of the studies (n = 9) used data directly obtained from clinical trials rather than a model-based approach for their economic analysis. The time horizon employed in the economic analysis spanned from 14 days to lifetime. Most studies (n = 12) received funding from national level granting agencies and industries, whereas two studies [32,33] received no funding. CEA: cost-effectiveness analysis; CUA: cost-utility analysis; CBA: cost-benefit analysis; CMA: cost-minimization analysis. Table 3 presents the summary of the 14 included studies. Of these 14 studies, pharmacists used a telephone-based intervention in seven. Five out of the seven studies reported the economic benefits of the phone interventions [32,33,35,36,38]. Dehmar et al.'s CEA study revealed that home blood pressure monitoring and pharmacist case management improved hypertension care in a cost-effective manner [35]. They reported that this intervention cost USD 7337 per person achieving hypertension control. Another CEA study by Faleh AlMutairi et al. found telemedicine-based care to be cost-effective for patients with type 2 diabetes [32]. They reported an incremental cost-effectiveness ratio (ICER) of SAR 2373 per 1% reduction in HbA1c. In a CBA study with polymedicated elderly patients admitted to an internal medicine ward, a clinical pharmacist carried out phone-based follow-up interventions after patients' discharge [36]. The interventions achieved a net cost saving of EUR 1301 by reducing readmissions of these patients. The net cost savings became greater in those with intermediate and high risks of potentially avoidable readmission. A CMA study was conducted using a pre-and post-design in which patients with HIV received pharmacist teleconsultations [33]. The economic evaluation of the teleconsultation interventions showed EUR 137 patient/year costs-saved and 18.5 h/patient/year working time gained. In Padwal et al.'s CUA study, patients with cerebrovascular condition were provided home blood pressure telemonitoring services with pharmacist case management interventions [38]. Their results using a Markov model demonstrated that telemonitoring with case management was dominant (i.e., less costly and more effective) compared with usual care. In contrast to these five studies, telemedicine-based care was not consistently considered to be cost-effective in two studies [19,20]. In their CUA study, Painter et al. included rural veterans with posttraumatic stress disorder (PTSD) assigned to either telemedicine care or usual care [19]. Although the telemedicine care intervention was more effective than usual care with an incremental increase in QALY of 0.008, the intervention was relatively expensive with an incremental cost of USD 2495, resulting in an ICER of USD 185,565/QALY. Therefore, the improvement in QALY did not translate to economic benefit due to the high cost of this intervention given the conventionally reported willingness-to-pay (WTP) thresholds of USD 50,000-USD 150,000/QALY. Of note, their subgroup analyses focusing on patient subgroups with anxiety disorder, depression, and panic disorder showed that the intervention was dominant (i.e., produced greater QALY at lower costs) (data not shown in Table 3). In Pyne et al.'s CUA study, patients with depression were randomly assigned to receive either telemedicine care or usual care [20]. In their base-case analysis including outpatient costs alone, the incremental intervention effects on expenditures (β = USD 1528, p < 0.001) and QALY (β = 0.018, p = 0.04) were significant, resulting in an ICER of USD 85,634/QALY. The ICERs ranged from USD 111,999 to USD 132,175/QALY in their secondary analysis, where inpatients costs were also included in addition to outpatient costs. The ICERs exceeded a WTP threshold of USD 50,000/QALY, indicating that telemedicine care might not be cost-effective given this threshold.

Computer-or Web-Based Intervention
Three CEA studies performed an economic evaluation of computer-or web-based interventions by pharmacists [12][13][14]. All these studies found the interventions to be cost-effective. In Avery et al.'s trial, patients given electronically prescribed prescriptions by a general practitioner were assigned to either the control group who received computer-generated simple feedback or the intervention group who received a pharmacistled intervention composed of feedback, educational outreach, and dedicated support [13]. The results of the study showed that the ICER of the intervention was GBP 66 per error avoided, indicating the intervention was cost-effective. In another trial by Fishman et al., patients with hypertension were randomized to receive either usual care plus a home blood pressure (BP) monitor or a pharmacist's web-based management intervention plus a home BP monitor [12]. The intervention was found to be cost-effective with ICERs of USD 2220 and USD 1850 per life year for women and men, respectively. In Hope et al.'s study, a tiered review method (computer-based review prior to a clinician's review) was compared with the traditional pharmacist-based review approach, both of which were used to identify adverse drug events (ADEs) and medication errors (MEs) in adults with outpatient care at ambulatory clinics [14]. This tiered review turned out to be more cost-effective than the traditional review (USD 42 vs. USD 69 per ADE identified, respectively).

Videotape-Based Intervention
Bosmans et al. conducted a randomized clinical trial in which subjects comprised those who had visited a pharmacy with a new prescription for a non-tricyclic antidepressant [34]. The subjects were randomly assigned to either the coaching program (consisting of three contacts with the pharmacist and a take-home video) or usual care. Outcome measures were treatment adherence and improvements in depressive symptoms measured using the Hopkins Symptom Checklist (SCL). Although the coaching program cost slightly more than usual care (EUR 3275 vs. EUR 2961), there were no significant differences in outcomes. Therefore, the authors concluded that the program was not cost-effective compared with usual care.

Smartphone-Based Intervention
Lowres et al. evaluated the economic value of pharmacy-based screening for unknown atrial fibrillation (AF) using an iPhone electrocardiogram (iECG) [37]. Compared with no screening strategy, screening with iECG was found to be cost-effective, with the ICERs of AUD 5988 per QALY and AUD 30,481 for preventing one stroke.

Multiple Techologies-Based Intervention
Two CEA studies conducted economic evaluations of multiple technologies-based interventions [15,16]. A clinical trial by Amador-Fernandez et al. included patients who presented with minor ailments or requested a non-prescription drug for such ailments [15]. The patients were randomized to receive either the intervention (comprising face-to-face consultation utilizing a web-based program plus telephone follow-up) or usual care. The intervention was considered to be cost-effective with an ICER of EUR 24,733/QALY given a WTP threshold of EUR 25,000/QALY. Those with minor ailments were also recruited in Dineen-Griffin et al.'s study [16]. The intervention group received face-to-face consultation using the technology-integrated platform plus telephone follow-up, whereas the control group received usual care alone. The intervention was found to be cost-effective with an ICER of AUD 2277/QALY. Table 4 summarizes the results of quality evaluation using the QHES instrument (actual points in Table 4 are available upon request). Overall, the quality of the included studies was good (57%) or fair (43%). No study was rated as poor.

Discussion
Despite growing interest among pharmacists in using digital technology in practice, there has been a paucity of research about the economic evaluations of digital interventions. Accordingly, the purpose of the current study was to systematically review the economic value of pharmacist-led digital interventions. Pharmacists used telephones with call functions only, computers, web-based interventions, videotapes, smartphones, or multiple technologies (such as web plus telephone or technology platform plus telephone) for their digital interventions. First, telephone-based interventions were generally found to be costeffective in the studies included in this review. That is, five out of the seven phone-based studies revealed the economic benefits of the phone interventions [32,33,35,36,38]. In the other two telephone-based studies by Pyne et al. and Painter et al. [19,20], the authors concluded that the interventions were not cost-effective because the ICERs of the interventions were higher than the conventionally cited WTP thresholds of USD 50,000-USD 150,000/QALY. However, these thresholds were proposed more than several decades ago. If these studies were reevaluated using more recently suggested thresholds such as USD 200,000/QALY or USD 300,000/QALY [39,40], these interventions could be economically favorable. Specifically, all ICERs reported in Pyne et al.'s study fall below the thresholds of USD 200,000-USD 300,000/QALY, which makes the intervention cost-effective. Similarly, the ICER from base-case analysis in Painter et al.'s study is lower than the threshold of USD 200,000-USD 300,000/QALY. Additionally, their subgroup analyses showed cost-savings of the intervention. That is, this intervention resulted in greater QALYs at lower costs among those with anxiety disorder, depression, or panic disorder. As such, Painter et al.'s findings suggest that phone intervention could be economically sound for individuals with comorbid mental disorders. This implies the importance of targeting specific population groups. It is crucial that clinicians, researchers, and policy makers identify population groups who can benefit most from digital interventions and consider prioritizing their needs in providing such interventions. Next, a pharmacist's computer-or web-based interventions were consistently found to be cost-effective in earlier studies [12][13][14]. However, in Bosmans et al.'s study, providing videotape-based intervention to patients with depression was found not cost-effective [34]. Because there were no significant differences in treatment adherence and depressive symptoms between the intervention and usual care groups, the intervention was considered an economically less preferred option. Of note, this study adopted a 6-month time horizon, and thus, the long-term effects of this intervention were not evaluated. Moreover, as the authors stated, their results were not consistent with those of two other studies reporting significant improvements in adherence to antidepressants after pharmacist interventions [41,42]. According to Bosmans et al., their intervention was not intensive-i.e., they deliberately delivered a minimal easy-to-implement intervention intended not to interrupt the pharmacist's daily practice and to minimize participant attrition. This could possibly explain the non-significant difference in adherence between the intervention and usual care groups. In the two other studies that showed significantly improved adherence in the intervention group, pharmacist interventions were more intensive. Perhaps, making the interventions in Bosmans et al.'s study intensive enough to improve treatment adherence would have led to favorable cost-effectiveness results. Therefore, the cost-effectiveness results of the pharmacist's interventions may be dependent upon how the interventions are designed. The degree of the pharmacist's engagement is likely to affect patient outcomes, which in turn influences the economic value of the intervention. Even though pharmacists provide digital interventions, if they are engaged to a limited extent in using technologies, the interventions may not improve patient outcomes as effectively as expected. This highlights the importance of pharmacists' dynamic use of digital tools to enhance patient outcomes. In addition, the smartphone-based intervention was considered cost-effective in a prior study [37]. Finally, two previous studies using multiple technologies revealed that interventions based on such technologies were cost-effective [15,16].
There may be several possible reasons for why technology-based interventions are cost-effective. First, patients who receive the intervention can save direct nonmedical costs (e.g., transportation or childcare costs) or indirect costs (e.g., productivity loss), for example, by receiving care in a remote location rather than in costly care facilities. Alternatively, early identification/detection of a disease using technology can prevent disease progression and costly health outcomes. For example, in one of the studies included in our review, the authors screened the elderly to identify new AF using iECG [37]. Those identified could have received prompt treatment to prevent stroke and to thus avoid the high cost burden of stroke. Such economic benefits can be obtained not only from preventing stroke but also from a wide range of conditions using a technology-based screening tool. Additionally, technology may enhance care providers' efficiency. Using the technologies, clinicians can decrease their time for each patient. This will ultimately lead to reduction in costs to clinicians themselves as well as to society.
The current study found that telephoning was the most frequently used intervention tool among pharmacists. This finding is consistent with the results from other studies [9,10]. Nine out of the 14 studies included in this review (64.3%) adopted a telephone-either as a single digital intervention tool or combined with another digital tool. Similarly, a telephone was used most frequently as a digital intervention tool in 69.2% of the community pharmacists' interventions [9] and 78.9% of the clinical pharmacists' interventions [10]. Notably, computers or web-based interventions were used as intervention tools in only three studies. A smartphone was selected only in one study. Other tools such as a wearable device or social media were not adopted in the studies we reviewed. As noted in our earlier study [10], the telephone represents the traditional method of digital interventions. Pharmacists can consider more novel and diverse technologies in their interventions to serve expanded purposes. For example, smartphone technologies have been used for smoking cessation [43,44], weight management [45], and treatment of eating disorders [46]. In addition, wearable sensors integrated with smartphone apps record an individual's real-time vital signs and health-related parameters, demonstrating the potential for personalized care solutions [47,48]. Pharmacists using these technologies to promote healthy behavior and better disease management, as noted in the examples, should result in positive outcomes in their patients. Once identifying how much these technologies can improve patient outcomes, the economic value of the interventions based on the technologies should be evaluated.
There have been challenges in conducting economic analyses in prior studies. Although pharmacist-led digital interventions were found to be generally cost-effective in the studies, these results were limited to short-term economic values in most cases. A total of 11 studies (78.6%) chose a very short time frame (e.g., a time horizon ≤ 12 months), and the economic value was evaluated with this time frame. Therefore, the long-term economic value of digital interventions remains unclear. Furthermore, in most studies, the economic value was evaluated without using an economic model-based approach. One of the benefits of using an economic model is estimating the future costs and outcomes, thereby facilitating the evaluation of long-term economic value. These prior studies did not likely use an economic model because they adopted such a short time frame. Moreover, some studies did not include direct nonmedical and/or indirect costs in their analyses. However, technology can significantly save time and resources and thus reduce such direct nonmedical and indirect costs. In this sense, economic benefits of digital interventions could be estimated more accurately by including these costs with a societal perspective. In future studies to evaluate the economic value of digital interventions, researchers need to consider using an economic model (e.g., a Markov model) incorporating a sufficiently long time frame and perspective by including the comprehensive types of costs.
Guidelines for evaluating internal and external validity in systematic reviews have been proposed by Avellar et al. [49]. According to the guidelines, authors of systematic reviews can report the important elements of internal and external validity such as information related to effectiveness, generalizability, applicability, and feasibility of an intervention.
Thus, we abstracted such information from the studies and presented them as follows. First, this review included a wide range of population groups such as adults with minor ailments, diabetes, hypertension, HIV, cerebrovascular conditions, depression, elderly using ≥ 2 medications, and veterans with PTSD. Because a summary measure in the economic evaluations is an ICER, we were not able to calculate numerical impact estimates (e.g., regression coefficients, odds ratio, and adjusted mean differences), effect sizes, or p-values of the impact estimates. Additionally, interventions included in our review were pharmacist-led digital interventions using telephones, computers, webs, videotapes, smartphones, and multiple technologies. In general, pharmacists needed experience and education to implement such digital interventions. The comparison group in the studies usually received usual care or traditional care. All the information here should be considered in generalizing the study findings to a broader situation. This is the first systematic review of the cost-effectiveness of pharmacist-led digital interventions. Our review provides a narrative integrating the economic outcomes for multiple digital technologies that may inform decision-making for clinicians and policy makers. However, our study has several limitations. First, we attempted to capture all relevant articles by following the recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. Nevertheless, this approach may have missed some articles. Second, it was not possible to quantitatively pool the outcomes from different studies because of heterogeneity inherent to the studies (e.g., study population, intervention components, and outcome measures). As such, we decided to present the findings of the study qualitatively. Furthermore, because our review relied on published articles alone, there may be unpublished findings that were not included in this review. This may raise the risk of publication bias. In addition, three studies included in this review [14,16,32] employed a time frame of less than 6 months, which could arguably be considered as pilot studies. Accordingly, these three studies could be excluded in the review in accordance with our inclusion/exclusion criteria. However, exclusion of these three studies did not affect our findings that pharmacist-led digital interventions were cost-effective. That is, digital interventions based on telephone, computer, and multiple technologies were still considered to be cost-effective even when these three studies were excluded in our review. Finally, this review included 14 studies only. Given the increasing attention in digital health, many studies are expected to be performed in the future to investigate the economic value of a diverse range of pharmacist-led digital interventions. As such, future literature reviews including a larger number of studies are necessary to confirm cost-effectiveness of such interventions.

Conclusions
This review found evidence supporting the short-term cost-effectiveness of pharmacistled digital interventions. Compared with traditional usual care, all interventions using telephones, computers, web-based interventions, or smartphones yielded more economic benefits. A video-based intervention was found cost-ineffective possibly because the intervention was not intensive enough to improve treatment adherence. This finding highlights the importance of aligning the design of an intervention with its purpose.
Economic evaluation of digital interventions can contribute to evidence-based clinical decision making. Clinicians can use evidence from such economic evaluations to decide whether to adopt technology tools in their practice. Additionally, such evidence can be considered in a value-based reimbursement system since economic value is one of the important determinants that influence access to care.
Funding: This research received no external funding.