Design
This study consists of two phases (Figure 1):
Phase 1: To determine the prevalence of dosing inadequacy, an observational, multicentre, cross-sectional study was carried out.
Phase 2: To evaluate the effectiveness of the community pharmacist intervention, a non-randomised controlled study with historical control group was carried out. The Ethics Committee at Germans Trias i Pujol Hospital revised and approved the study (EO-12-038).
Setting and participants
The study was carried out by 40 volunteer pharmacists in 40 community pharmacies, accredited by the University of Barcelona for training senior pharmacy students, from various areas of Barcelona (Spain): Barcelonès nord, Maresme, Vallès occidental, Vallès oriental, Berguedà and Osona. The participant pharmacists received two sessions (8 hours) of training prior to the study in which they were trained to detect a patient with renal impairment and make a dosage adjustment if needed.
Patients over 65 years old that presented to one of the participant community pharmacies with 3 or more prescriptions were invited to participate in the trial by the pharmacists (Phase 1: Oct 2010-Jan 2011; Phase 2: Feb 2011-May 2011). The inclusion criteria were: people over 65 years old; taking 3 or more drug products; having a Body Mass Index (BMI) between 19 and 35 Kg/m2; not being vegetarian or on a high protein diet; not having had a limb amputated, paralysis or muscular diseases; not taking creatinine-affecting drugs. Written informed consent was obtained for all participants.
Participants were screened by pharmacists (a minimum of 25 participants screened per pharmacist) for renal impairment by measuring the glomerular filtration rate (GFR) and creatinine clearance (CrCl). Serum creatinine values were used to calculate the CrCl (Cockcroft-Gault method) and the glomerular filtration rate (Modification of Diet in Renal Disease MDRD-4 method) with the help of an online renal calculator [19]. Patients with GFR levels lower than 60 ml/min/1.73 m2 and/or a CrCl lower than 60 ml/min were considered as having a renal impairment and were included in the study (Figure 1).
Variables
Data on variables were collected by the pharmacist in an interview with the patient at the community pharmacy.
Sociodemographic variables: age and gender; anthropometric variables: Height (m), measured through calibrated height scales; weight (Kg): measured through calibrated weight scales; Body Mass Index (BMI): in kg/m2; Body surface area (m2) calculated through Mosteller’s method [19]. Clinical Variables: Blood Pressure (BP) (mmHg), measured through calibrated and validated instruments; Serum creatinine (SCr) in mg/dL through Reflotron® or blood test conducted in the previous 3 months at the Primary Care Health Centre; comorbidity (hypertension, diabetes, dyslipidemia, hyper/hypo thyroidism, cardiovascular disease (CVD) reported by the patient). Pharmacological treatment: anatomical and therapeutic and chemical group (ATC), active ingredient, doses, dosage, and dosing interval.
Intervention
Certain drug databases, such as CIMA [20] (Medicines Online Information Centre) the Spanish official drugs Information, Martindale [21] and American Hospital Formulary Drug System Information [22] (AHFS), were consulted to study the dosing inadequacy of the active ingredients contained in the drug products that the patients were taking. When an active ingredient requiring dose adjustment (Cause of DRP Code C3.2, C3.4 of Classification of Pharmaceutical Care Network Europe Foundation PCNE) [9] was detected or when it was contra-indicated (Code C1.1 of PCNE) [9], we considered the patient as having a potential DRP (PCNE Code P2.1, P2.3) [9].
The prevalence of dosing inadequacy (DI) was calculated through the mathematical expression [12]: DI = (N° inadequately adjusted and/ or contra-indicated drugs / N° total adjustable drugs)*100
Pharmacists used a questionnaire prepared by the research team to write a brief report to the general practitioner (GP) detailing the problems that had been detected and suggesting changes in pharmacological treatment (change drug, decrease quantity of dosage, lengthen time interval between doses). In this written report there was a space that allowed the GPs to provide the pharmacists with a written answer. These written reports were delivered to the GPs by the pharmacists (face-to-face or using the GP’s mailboxes). After 7–14 days the pharmacists went to the Primary Care Health Centre (PCHC) to collect any replies from the GPs. The reports included the pharmacists’ telephone numbers to allow the GPs to contact them.
Sample size calculation
In Phase 1 the objective was to determine the prevalence of dosing inadequacy in the elderly polymedicated population with renal impairment. Assuming that the prevalence of dosing inadequacy in the elderly is 20%, as has been previously reported [12], and to obtain a precision of 3% with a 95% confidence interval, the sample size required was 683 drug products with restrictions in case of renal impairment.
In the second phase, the aim was to evaluate the effectiveness of the community pharmacist intervention in addressing the problem of dosing inadequacy when compared with usual care. Taking into account expected dosing inadequacy values of 20% in the control group and 10% in the intervention group and to detect differences between groups in the levels of dosing inadequacy with a power of 95% and a 95% confidence interval, a minimum of 328 drug products with restrictions in case of renal impairment per group was required.
The number of drug products needed to assess the prevalence of requirements for dosing adjustments was higher than the number of drugs needed to evaluate the effectiveness of the pharmacist intervention. Consequently, 176 patients from the 263 that had been included in Phase 1 formed a randomly chosen subsample to be used as historical control group in Phase 2 (Figure 1).
Statistical analysis
Data analysis was performed using the Statistical Package for the Social Sciences version 19 (SPSS). Quantitative variables were expressed as means. Other variables were expressed as relative and absolute frequencies. The comparison between variables was conducted using the Chi-Squared test and Student’s t-test for qualitative and quantitative variables, respectively. ANOVA was also used to compare qualitative variables for multiple layer quantitative analyses. P values less than 0.05 were considered statistically significant. Population parameter estimates were carried out at a Confidence Interval (CI) of 95%.