Skip to main content

Table 1 Characteristics of studies included in this systematic review

From: Does drug dispensing improve the health outcomes of patients attending community pharmacies? A systematic review

Study

Country

Design

Sample size

Evaluated outcomes

Main results

Limitations and bias described

Chong et al., 2011 [49]

Australia

Cross-sectional

97 community pharmacists

Cost-savings for patients who accepted a generic substitution.

The average cost-saving per item was AUD$2.26.

Low response rate; survey was time-consuming; pharmacist selection bias; limited representation of the collected brand name prescriptions; the findings cannot be generalized to whole population of community pharmacies.

Basheti et al. 2008 [50]

Australia

Single-blind cluster randomized parallel group design

112 patients

Peak expiratory flow (PEF) variability, asthma severity; asthma-related quality of life.

In the active group, asthma severity was significantly reduced at 2 months (p = 0.001), 3 months (p < 0.001), and 6 months (p = 0.015) compared with the control group. By 3 months, only 8 % of active group patients were classified as having severe asthma compared with 22 % of control group patients (p = 0.037). Post-hoc analysis demonstrated a significant correlation between improvement in Inhaler Technique Score and improvement in both PEF variability (r = 0.31, p = 0.008) and asthma related quality of life( r = 0.37, p = 0.001).

Not reported.

Crockett et al. 2006 [51]

Australia

Parallel group design with a control and intervention group

106 patients (60 in the control group and 46 in the intervention group)

Patient’s psychological wellbeing (K10 scores), patient satisfaction with service.

Wellbeing was improved in control and intervention groups (the K10 scores decreased significantly from baseline to 2 months in the control and intervention groups by 4 points [p < 0.0001] and 4.7 points [p ≤ 0.001], respectively). However, the K10 score for the intervention group continued to decrease in the second month whereas that of the control group increased slightly (16.7/50 versus 17.7/50 control [p = 0.176]; 20.5/50 versus 18.3/50 [p = 0.009] intervention).

The project was of insufficient duration to effectively measure the full effect of the pharmacists’ intervention; telephone contact and K10 administration might have had beneficial effects and masked the impact of pharmacists in the intervention group; problems were experienced with the video-conference link, which might have limited the efficacy of training.

Ali et al. 2019 [52]

United Arab Emirates

Cross-sectional

210 patients

Patients’ perception towards pharmacist’ performance and satisfaction with dispensing related to privacity.

A total of 136 (64.7 %) patients strongly agreed or agreed that they were satisfied with the pharmacist counselling regarding the questions asked before dispensing medications, such as history of previous drug allergy, disease details, etc. 193 (91.9 %) patients disagreed or strongly disagreed that they were satisfied with the privacy maintained by pharmacist while discussing with patients and dispensing medications.

The results of this study cannot be generalized to all patients in Dubai; limited numbers of participants; closed-ended questions in the questionnaire may not help to clarify expectations.

Merks et al. 2019 [53]

Poland

Randomized trial

199 patients (102 in control group and 97 in intervention group)

Relief of symptoms after antibiotic therapy.

A total of 89 patients (91.7 %) in the intervention group (pharmacy practice with pictograms) and 86 patients (84.3 %) in the control group (usual pharmacy practice) reported symptom relief after antibiotic therapy (OR: 2.07, 95 % CI: 0.84–5.08, p = 0.1127). Regarding the subjective assessment of patients’ perspective, the net promotor score was higher for the intervention group (71.3 %, n = 69) than for the control group (51.5 %, n = 52). The chance that a patient was an advocate of pharmaceutical services was twice as likely in the case of pharmaceutical practice supported by pictograms (OR 2.04; 95 % CI 1.09 3.81; p = 0.0239).

Limited number of patients; authorial questionnaires were only validated in terms of face and content validity; subjectivity in outcome selection; psychological bias in patients’ responses (due to the interviewers being pharmacists); participation bias due to the randomized cluster design (patient randomization was not possible); duration of the study was relatively short.

O’Dwyer et al. 2020 [54]

Ireland

Cluster randomized open-label clinical trial

152 patients (74 in the biofeedback group, 56 in the demonstration group, and 22 in the control group)

Quality of life; self-reported respiratory symptom (presence of cough, breathlessness, and nighttime symptoms); asthma exacerbation rate.

In both the biofeedback and demonstration groups, there were statistically significant reductions in the total quality of life scores from month 1 to month 2, with decreases of 5.3 and 5.7, respectively. However, only patients in the biofeedback group had a sustained reduction to month 6 (-6.1; 95 % CI, -9.68 to -0.4; p = 0.04). There was a decrease in daily respiratory symptoms in the biofeedback group. Over the 6-month study period there was a numerically smaller number of moderate exacerbations of airway disease treated with both antibiotics and oral corticosteroids in the biofeedback group, 0.7 (95 % CI, 0.4–1.1) compared with 1.1 (95 % CI, 0.5–1.7) in the demonstration group, and 0.9 (95 % CI, 0.1–1.7) in the control group.

The cluster randomized design can allow significant imbalances between groups to occur at baseline. There was an interval of 3 months where all study procedures ceased. Recruitment was only 80 % of original target. Small number of patients per cluster and high dropout rate in the second half of the study.

Westerlund, 2009 [55]

Sweden

Cross-sectionala

89 community pharmacies

Economic outcomes of community pharmacy interventions in patients’ drug-related problems (DRP).

Pharmacy interventions have saved 68 (13 %) primary care contacts and 16 (3 %) future hospitalizations. The potential societal cost savings extrapolated to Sweden on the national level were estimated to be €358 million.

DRP were assessed and interventions were documented by one pharmacist and one physician; a relatively small number of DRP and interventions were evaluated; extrapolation to the national level in Sweden should be interpreted as demonstrating the potential societal cost savings by pharmacy interventions rather than as showing actual cost savings.

Payne et al., 2019 [56]

United States of America

Cross-sectionala

200 patients

Total money saved with dispensing services for each averted adverse event.

The estimated cost savings for each adverse event avoided resulted in a minimum cost savings of $10,458.

The study was performed in one independent pharmacy. The intervention patterns in this pharmacy are not generalizable; a retrospective data analysis was performed, making it unclear whether the annotations resulted in further activities that increased patient safety; the exact clinical outcome of each individual activity is unknown; data were collected over a 2-month period.

Ferreira et al., 2018 [57]

Brazil

Pre-post-test design (quasi-experimentala)

104 patients

Patient satisfaction with service.

Patients reported a high level of satisfaction towards the dispensing service, which was rated excellent or very good by more than 70 % of the patients.

The Morisky scale used in the study tends to overestimate non-adherence behavior; satisfaction may be overestimated, particularly considering that the patients interviewed were frequent users of the University Pharmacy.

  1. AUD Australian dollar, CI confidence interval, DRP drug-related problems, Or odds ration, PEF peak expiratory flow
  2. aClassification of study design by the authors of this systematic review