Table 3.

Results of included studies (N = 6)^a

Evaluation	Community Pharmacy Medicines Management Project Evaluation Team12	Dierick-van Daele et al13	Lee et al14	Neilson et al15	Richardson et al16	Turner et al17
Year of publication and country	2007, England	2010, Netherlands	2004, South Korea	2015, UK	2013, England	2008, UK
Intervention groups compared; size, type of role substitution, and setting	Intervention: pharmacist (n= 62). Control: GPs (n= 164); pharmacist-led medicines management versus standard care from the GP in nine general practices	Intervention: nurse practitioners (n= 12). Control: GPs (n= 15); role substitution of nurse practitioners by GPs in 15 general practices	Intervention: CHP (n= 600). Control: care delivered by physician; CHP services versus no-CHP services in primary health care. Postal survey questionnaire sent to a sample of CHPs nationwide	Intervention 1: pharmacists medication review with pharmacist prescribing (n= 70). Intervention 2: pharmacist review only (n= 63). Control: TAU from GP (n = 63); pharmacy-led care versus TAU for the management of chronic pain in six general practices	Intervention 1: nurse-led PR (n= 85). Intervention 2: nurse-led SL (n= 97). Control: TAU from GP (n= 92); nurse-led supported self-management compared with TAU in 186 general practices	Intervention: nurse-led care (n= 505). Control: TAU from GP (n= 658); nurse-led disease management versus standard GP care in 20 general practices
Type of economic evaluation analysis	Cost-minimisation	Cost-minimisation	Cost-minimisation	Cost-utility	Cost-effectiveness	Cost-utility
Main outcomes measured, type of costs measured, type of outcomes measured	Total NHS costs; direct costs of delivering the intervention and NHS treatment costs (for example, cost of medicines), and indirect costs of training (for example, attendance fees); and appropriate treatment and health status (SF-36 and EQ-5D)	Costs of GP versus nurse practitioner consultation; direct costs within the healthcare sector and costs outside of the healthcare sector (productivity losses); and process outcomes and outcomes of care	Total costs of care between CHP services model and no-CHP services model of care; direct costs (for example, personnel costs, materials) and indirect costs (operational and depreciation costs) of CHP services. Direct costs (for example, outpatient costs) and indirect costs (travel and loss of earnings) of no-CHP services; outcomes not assessed. The efficacy of the intervention was based on previous findings	Differences in mean total costs and effects of pharmacist-led management versus GP-led management of chronic pain; direct costs, other costs borne by patients, and productivity losses; and health utility derived from SF-6D	QALYs derived from the EQ-5D; cost to the NHS (for example, resource use and unit costs), private expenditures, informal care costs, and loss of production costs; health utility measured using EQ-5D	QALYs derived from the EQ-5D; direct costs (including travel costs); health utility measured with EQ-5D
Perspective of analysis	Not stated	Practice and societal	Not stated	NHS	NHS and personal social services	Patient and NHS
Currency and cost year	GBP derived from general practice-held records. Cost year not reported	Euros derived using the price index of Statistics Netherlands for cost year 2006	South Korean Won converted to USD derived from national unit costs for cost year 1999	GBP derived from PSSRU and British National Formulary for prices at cost year 2009/2010	GBP derived from NHS reference costs and PSSRU at 2008/2009 prices	GBP derived from HRGs for cost year 2002 to 2003 and inflated to 2003/2004 prices
Discounting and sensitivity analysis	Follow-up period 12 months, no discounting, no sensitivity analysis	Sensitivity analysis varying GP salary, no discounting	6-month time horizon, no discounting, no sensitivity analysis	6-month time horizon, no discounting. Three sensitivity analyses were conducted with imputed values for SF-36 scores	Costs and outcomes were discounted at a rate of 3.5% per year. A complete case analysis as part of sensitivity analysis was conducted	Follow-up period 12 months, no discounting, discount rate of 6% for equipment and training that would have an expected lifespan of more than 1 year. No sensitivity analysis
Intervention costs and main findings	Total NHS costs at baseline: intervention group, 852 GBP; control group, 738 GBP. Total NHS costs at follow-up: intervention, 971 GBP; control, 835 GBP. Statistically significant difference (P<0.01) in total NHS costs, due to the costs of providing pharmacists’ training. Mean difference in costs of 135 GBP (164 GBP inflated to 2016/2017 prices)18	Cost per NP consultation, 32 Euros; cost per GP consultation, 40 Euros; lower direct consultation costs for NP compared with GP (P= 0.01) mean difference 8 Euros (7 GBP inflated to 2016/2017 prices)18	Mean direct costs: CHP, 2424 USD (SD 566) (2520 GBP inflated to 2016/2017 prices).18 Physician, 5188 USD (SD 3262.5 USD) (5394 GBP inflated to 2016/2017 prices).18 Mean indirect costs CHP (500 USD, SD 258 USD) physician (1269 USD, SD: 952 USD); t-test found a significant difference in the average costs of care between the groups (P= 0.01 cost ratio of 2.16, with a range of 0.09 to 9.63)	Unadjusted total mean costs: prescribing group 452 GBP (509 GBP in 2016/2017 prices18); medication review group 570 GBP (642 GBP in 2016/2017 prices18); TAU group 1333 GBP (1500 GBP in 2016/2017).18 After adjusting for baseline costs, TAU was more costly than both pharmacy-led interventions (P= 0.01); cost differences relative to TAU was 78 GBP (87 GBP inflated to 2016/2017 prices)18 for prescribing group and 54 GBP (61 GBP inflated to 2016/2017 prices).18 However, once ICERs were calculated, both pharmacy-led interventions were more costly, with slightly higher QALY gains than TAU	Excluding intervention costs of SL and PR, at 70-week follow-up, total NHS cost of chronic fatigue syndrome 789 GBP for PR, 916 GBP for SL, and 710 GBP for TAU; TAU was slightly more effective than PR and SL, at a lower cost, when baseline differences in EQ-5D were adjusted. The nurse-led PR intervention produced a cost per QALY of 39 583 GBP (inflated to 44 812 GBP in 2016/2017)18	Total mean NHS delivery costs nurse 1108 GBP and GP 661 GBP (P<0.01); a difference of 0.03 QALY value was reported between the nurse-led group and usual care, and the cost per QALY gained in the nurse-led group was 13 158 GBP (17 694 GBP inflated to 2016/2017 prices)18
Conclusions	No change in numbers of patients receiving appropriate treatment. Pharmacy-led group more costly than standard care	Direct costs of consultations were lower for NPs than GPs. The differences in costs were mainly due to differences in salary	Care provided by a physician was twice as costly as the CHP services due to travel costs and loss of earnings for patients who would have had to travel to inner-city clinics to see a physician	The authors concluded that pharmacy-led management is more costly than usual treatment and produces similar QALYs compared with TAU	Nurse-led PR intervention was not cost-effective	Nurse-led disease management programme was cost-effective

^aCosts are given with no decimal places and CIs and P-values to two decimal places. CHP = community health practitioner. CI = confidence interval. EQ-5D = European Quality of Life-5 Dimensions measure. HRG = healthcare resource groups. n = number of individuals. ICER = incremental cost-effectiveness ratio. NP = nurse practitioner. PR = pragmatic rehabilitation. PSSRU = Personal Social Services Research Unit. SF-36 = Short Form-36 Health Survey. SF-6D = Short Form-Six Dimension health index. SL= supportive listening. TAU = treatment as usual. QALY = quality-adjusted life years.