Skip to main content

What is in your wallet? A cluster randomized trial of the effects of showing comparative patient out-of-pocket costs on primary care prescribing for uncomplicated hypertension



Drug expenditures are responsible for an increasing proportion of health costs, accounting for $1.1 trillion in annual expenditure worldwide. As hundreds of billions of dollars are being spent each year on overtreatment with prescribed medications that are either unnecessary or are in excess of lowest cost-effective therapy, programs are needed that optimize fiscally appropriate use. We evaluated whether providing physicians with information on the patient out-of-pocket payment consequences of prescribing decisions that were in excess of lowest cost-effective therapy would alter prescribing decisions using the treatment of uncomplicated hypertension as an exemplar.


A single-blind cluster randomized trial was conducted over a 60-month follow-up period in 76 primary care physicians in Quebec, Canada, and their patients with uncomplicated hypertension who were using the MOXXI integrated electronic health record for drug and health problem management. Physicians were randomized to an out-of-pocket expenditure module that provided alerts for comparative out-of-payment costs, thiazide diuretics as recommended first-line therapy, and tools to monitor blood pressure targets and medication compliance, or alternatively the basic MOXXI system. System software and prescription claims were used to analyze the impact of the intervention on treatment choice, adherence, and overall and out-of-pocket payment costs using generalized estimating equations.


Three thousand five-hundred ninety-two eligible patients with uncomplicated hypertension were enrolled, of whom 1261 (35.1%) were newly started (incident patient) on treatment during follow-up. There was a statistically significant increase in the prescription of diuretics in the newly treated intervention (26.6%) compared to control patients (19.8%) (RR 1.65, 95% CI 1.17 to 2.33). For patients already treated (prevalent patient), there was a statistically significant interaction between the intervention and patient age, with older patients being less likely to be switched to a diuretic. Among the incident patients, physicians with less than 15 years of experience were much more likely to prescribe a diuretic (OR 10.69; 95% CI 1.49 to 76.64) than physicians with 15 to 25 years (OR 0.67; 95%CI 0.25 to 1.78), or more than 25 years of experience (OR 1.80; 95% CI 1.23 to 2.65). There was no statistically significant effect of the intervention on adherence or out-of-pocket payment cost.


The provision of comparative information on patient out-of-pocket payments for treatment of uncomplicated hypertension had a statistically significant impact on increasing the initiation of diuretics in incident patients and switching to diuretics in younger prevalent patients. The impact of interventions to improve the cost-effectiveness of prescribing may be enhanced by also targeting patients with tools to participate in treatment decision-making and by providing physicians with comparative out-of-pocket information on all evidence-based alternatives that would enhance clinical decision-making.

Trial registration


Peer Review reports


Drug expenditures are responsible for an increasing proportion of health costs, accounting for $1.1 trillion in annual expenditure worldwide. Even with projected slowed growth from 9% in 2014 and 2015 to about 4–7% over the next 5 years, it will reach approximately $1.5 trillion by 2021 [1,2,3]. Health care systems in many countries face the same challenge—health system sustainability when cost increases outstrip economic growth [4]. Health systems need to provide adequate access to essential medication for their citizens in a fiscally responsible and sustainable manner. Programs are needed that can support optimal appropriate prescribing and discourage unwarranted drug expenditures. For example, in 2011, approximately $158 to $226 billion were spent in overtreatment in the USA alone [5,6,7,8,9].

Governments and third-party payers have led most of the initiatives to improve drug use and control drug costs [5, 7, 10, 11], primarily through consumer-directed user fees such as deductibles, co-payments, co-insurance, and/or spending caps [12,13,14]. User fees are intended to discourage the use of marginal therapies that have limited effectiveness in maintaining or improving health status [7, 12, 13, 15]. Systematic reviews of the effects of cost-sharing policies have shown that these interventions have led to reductions in medication adherence for both essential and less essential therapy [12, 13, 15,16,17], with a commensurate increase in medical visits, emergency care, and hospital admissions [13, 15,16,17]. For this reason, consumer-directed policies, such as cost-sharing, are seen to play a limited role in optimizing prescription drug use.

Physician-directed policies should be more successful in improving cost-effective use of medications since almost all prescribing decisions are made by physicians [15, 18]. Tiered formularies and reference-based pricing are two of the most commonly used policies that aim to influence the choice of medication prescribed towards more “cost-effective therapies” [19,20,21,22,23]. Both of these strategies use out-of-pocket payments by patients to influence physician treatment choices—where higher out-of-pocket payments are required for drugs that are considered to be less “cost-effective.” Several flaws with these popular approaches have been identified. Physician knowledge of the cost of drugs they prescribe is poor [24,25,26,27], even though most physicians consider cost to be an important factor in prescribing decisions [27, 28], particularly when it affects their patients’ out-of-pocket payments [29]. As a result, the pharmacists or third-party benefit managers then assume the responsibility for calling physicians to request changes in drugs prescribed to adhere to drug choice policies [25]. These call-backs to physicians are time and labor-intensive [25, 30, 31]. Patients frequently decide to lower their out-of-pocket expenses by discontinuing medication or to ration use of drugs with higher co-payments [21,22,23, 32,33,34,35,36,37]. These patient-directed solutions result in unintended consequences since the majority of medications are used for chronic disease management, and under-use of essential therapy is associated with an increased risk of avoidable morbidity [38,39,40,41,42,43].

Physician-directed cost containment could potentially be more effective if physicians knew at the time of prescribing their patients’ expected out-of-pocket expenditure [44, 45]. Computer-assisted prescribing and decision-support systems can be designed to display cost information at the time of prescribing, but there has been limited study of the potential value of this tactic in primary care where the majority of drugs are prescribed [45,46,47,48]. Of interest, although recommended [44, 45, 49], no intervention to date has specifically addressed one of the most important elements—physician’s knowledge of the consequences of prescribing decisions on the out-of-pocket costs for their patients. Observational studies suggest that physicians will prescribe lower-cost medication or use samples to reduce financial burden [29, 50], when it is the patient who has to pay rather than the insurance company [50]. However, it has been difficult to provide out-of-pocket payment information at the time of prescribing as factors involving the patient’s insurance plan, deductible, and co-payment requirements can be independently modified with little notice. With the recent implementation of electronic prescribing and integrated drug management systems at prototype centers in Quebec, it is now possible to develop and test the potential benefit of providing physicians with comparative out-of-pocket costs for patients at the time of prescribing.

We tested the hypothesis that the provision of comparative information on expected patient out-of-pocket expenditures to physicians at the time of prescribing will increase the proportion of patients started on or switched to more cost-effective treatment and improve medication adherence. The treatment of uncomplicated hypertension was selected as a test case because (1) multiple drugs exist to treat the same problem, (2) substantial differences exist in treatment costs among available choices, (3) a large proportion of the population is treated for the condition, (4) level-one evidence exists from randomized controlled trials showing that treatment among the available options is equivalent, and (5) a substantial proportion of the population are not receiving the most cost-effective therapy [51,52,53,54,55,56].



This study was conducted in the province of Quebec, Canada. The provincial health insurance agency, Régie de l’assurance-maladie du Québec (RAMQ), provides coverage for all eligible residents and pays all physicians and community pharmacists on a fee-for-service basis. All provincial residents are required to have drug insurance, either through enrollment in their employer’s benefit plans, or if not available, in the RAMQ public plan. Approximately 50% of residents are enrolled in the public plan, including all persons 65 years of age or older, welfare recipients, and persons who could not be insured through their employer. In the public plan, welfare recipients and seniors who receive a full guaranteed income supplement from the government are not required to share in the cost of their prescriptions. All other beneficiaries are required to share in the cost of their prescriptions which, during the period of this study, comprised a fixed monthly deductible of $16.50 and a co-pay of 25% of the cost of the drug plus the pharmacist’s dispensing fee. If a beneficiary’s out-of-pocket costs in a given month exceeded $52.65 for those receiving a partial government income supplement or $88.83 for those receiving no income supplement, the full costs of all additional prescriptions would be covered by the RAMQ plan.

Data generated by administrative claims for pharmacist and physician reimbursement in the RAMQ health and drug insurance program have been validated for clinical and research use [57,58,59,60,61]. In 2003, an experimental primary care-based electronic prescribing solution, the Medical Office of the XXIst century (MOXXI), was established and was the first system to connect to administrative health databases and integrate this information into a clinical electronic health record system [62].

Design and study population

A single-blind, cluster randomized controlled trial was conducted over a 60-month follow-up period to test the hypothesized benefits of providing comparative patient out-of-pocket expenditure information at the time of prescribing anti-hypertensive medication. Both newly treated (incident), defined as those patients having no anti-hypertensive prescriptions or dispensed medications in the prior 12-months, and currently treated (prevalent) patients, those with evidence for active therapy in the prior 12-months, were studied. The trial was conducted within an existing cohort of primary care physicians and their patients in the two urban centers of Montreal and Quebec City. We based our sample size on the ability to detect a minimum difference of 10% (an increase from 20 to 30%) in the proportion of patients with uncomplicated hypertension who are prescribed thiazides for hypertension management.

Physicians were eligible for inclusion if they were participants in the MOXXI electronic health record platform. This cohort, comprised of 76 full-time community-based primary care physicians in fee-for-service practice, represents 25% of all eligible primary care physicians within the geographically defined areas of the two cities. Patient inclusion criteria included age 18 years or older, insured by RAMQ and were required to share the cost of their prescriptions, made one or more visits to an enrolled study physician during follow-up, had a diagnosis of uncomplicated hypertension in the year before or at the visit (i.e., hypertension recorded by the study physician in the therapeutic indication field that must be completed with each prescription), and the study physician was the prescriber of their anti-hypertensive therapy. Patients were excluded if the following comorbidities were identified at the time of enrollment, as treatment regimens differ for complicated hypertension: diabetes, congestive heart failure, atherosclerotic disease, peripheral arterial disease, ischemic heart disease—angina or prior myocardial infarction, past cerebrovascular accident or transient ischemic attack, renal disease, asthma, chronic obstructive pulmonary disease (COPD), or left ventricular hypertrophy. Patient eligibility was determined at each visit during follow-up by an automated review of changes to the electronic problem list [61, 63]. Once a patient was considered eligible, they were followed until the completion of the study, death, admission to long-term care, or a move out of the province, whichever came first. All costs are reported in Canadian dollars.

Intervention and control group

Control group

The control group physicians had access to the basic computerized MOXXI medical record system but not the out-of-pocket expenditure module. MOXXI is a secure web-based electronic prescribing solution that provides a number of clinically useful features that increase the adoption and use of this technology in primary practice [63,64,65,66]. Information for each patient is retrieved in real-time from the RAMQ administrative databases, including their drug insurance coverage, dispensed prescriptions, and medical service claims. MOXXI presents drug information as a dynamic graphical display showing all currently active medications, color-coded by prescribing physician, as well as drug costs and dates of ER and inpatient hospital episodes based on medical service claims. A potential list of medical conditions and problems is generated from recorded treatment indications for drugs prescribed as well as by inclusion of ICD9 codes from medical service claims. This list is presented to physicians for validation. Prescription printouts are given to the patient and kept in the chart. To enhance efficiency, each patient’s electronic record is pre-populated with his or her demographic, drug, and medical visit information from the RAMQ claim data, and prescription refills are expedited by providing a “quick click” re-prescribing function for multiple repeat prescriptions. A commercial drug knowledge and alert system is integrated into the MOXXI system that provides physicians with drug monographs and medication review and alerts for potential drug interactions, therapy duplications, drug disease and allergy contraindications, and excessive doses. Alerts are classified into three levels of severity. The MOXXI system allows physicians to set the level of alerts they wish to see and eliminate alerts that they think are irrelevant for some or all patients. Most MOXXI physicians elect to see only the most severe alerts (absolutely contraindicated) that represent approximately 5% of all alerts [67].

Intervention group

A comparative out-of-pocket expenditure module was developed and incorporated into the MOXXI system to provide decision support to the prescribing physician (1) for selecting the most cost-effective drug for new anti-hypertensive treatment and (2) to facilitate switching patients who are currently treated for uncomplicated hypertension to more cost-effective therapy.

i) Decision support for newly treated hypertension patients: Decision-support recommendations were integrated into the electronic prescription pad. During typical use, a drop-down menu of drugs with their corresponding commonly prescribed dosages and frequencies appears as the physicians enter the first letters of the drug name. Once the physician finds and selects the desired drug sentence, the information is used to automatically populate the electronic prescription, which can be further modified if needed. If a physician is intending to prescribe a new anti-hypertensive agent for uncomplicated hypertension, a pop-up window will open automatically, showing alternate drug choices and the corresponding out-of-pocket and total costs of each treatment. Expected annual out-of-pocket costs are calculated based on the amount that the RAMQ will pay for the drug(s) (per tablet or capsule), the frequency selected, and the pharmacist’s dispensing fee. We assume the pharmacist will adhere to the recommended provincial practice of providing a 1-month supply (12 dispensings/year) [68]. We then apply the monthly deductible and 25% co-payment to the annual cost of the drug and pharmacist’s fees to determine annual out-of-pocket costs. The physician then has the option of selecting an alternative therapy and the system will automatically generate an electronic prescription for that selection (Fig. 1). The design of the intervention is based on prior findings that indicate that physicians want to be informed of more cost-effective option at the time of prescribing [25, 29] and are more likely to prescribe the more cost-effective option if the alternative option can be selected easily and quickly [69]. If physicians decided not to switch to a diuretic as first-line therapy, they were asked to document the reason for their decision.

Fig. 1
figure 1

Decision support for newly treated hypertension patients. Pop-up with patient cost information upon selection of the indication “Essential (primary) hypertension”

ii) Decision support for currently treated hypertension patients: We anticipated that physicians are more reluctant to switch existing treatment for patients who have already been started on a less cost-effective therapy, unless the patient is not achieving the desired treatment response, is not adherent to treatment (possibly because of cost), and/or has difficulty paying the out-of-pocket costs. A pop-up window appears when the anti-hypertensive is being renewed. This window provides information about the patient’s current annual out-of-pocket payment and the annual savings that would occur with a change in treatment (Fig. 2). Out-of-pocket costs were calculated using the same method outlined for newly treated patients. Physicians can click on the highlighted medication to view medication compliance within the last 6 months, drug details and cost, and can click a button to switch patients to the recommended therapy. The resulting changes in treatment are displayed in the prescription preview and printed as part of the prescription transmitted to the pharmacy. If physicians decided not to switch treatment, they were asked to document the reason for their decision and the alert would then be suppressed.

Fig. 2
figure 2

Decision support for currently treated hypertension patients. Pop-up with patient cost information upon selection of the indication “Essential (primary) hypertension”. Past adherence information is also available to allow a more informed decision

Randomization and blinding

Physicians were stratified by city, practice population size, and proportion of prevalent patients on diuretic treatment, with groupings sufficient to maintain a minimum of two physicians within each stratum. Within each stratum, we assigned a random number to each physician, and an equivalent number of physicians within each stratum were allocated to the intervention and control groups. Randomization took place 2 months prior to the start of the intervention, to permit sufficient time for physician training, and was carried out by an independent statistician, blinded to physician identity. Due to the nature of the intervention, it was not possible to blind either physicians or patients, but both were blinded to the primary and secondary outcomes of the study, as were the research assistants involved in training, implementation, and analysis, although physicians may have suspected that the primary outcome was an increase in diuretic use.

Physician training and support

All physicians enrolled in the MOXXI program receive one-on-one training in their clinic on how to use MOXXI. Their practice population information is pre-loaded prior to the start of training. The effectiveness of training is determined by the completion of a standardized set of tasks in prescribing, stopping, and modifying drug treatment, and adding problems to the problem list, at the end of the training and 1 month later. A local technical and training team is on call to address problems. The team proactively monitors utilization rates and intervenes if utilization is low. The training team provided one-on-one training on how to use the out-of-pocket training module in the clinic, monitored its use, and periodically visited each practice to provide proactive support.


Primary outcome: prescription of cost-effective treatment for uncomplicated hypertension

The intended effect of the intervention was to influence the physician’s choice of the drug prescribed to manage hypertension. We classified an individual’s anti-hypertensive drug treatment into one of three mutually exclusive categories: (1) thiazides alone, (2) other evidence-based but less cost-effective single drug therapy (angiotensin-converting enzyme inhibitors, beta blockers, calcium channel blockers, angiotensin 2 receptor blockers), and (3) other non-evidence-based choices including concurrent multiple anti-hypertensive drugs. These categories were derived from the 2016 Canadian Hypertension Education Program Clinical Guidelines [70]. Patients were classified according to drugs that were prescribed on the first visit during the follow-up period based on data retrieved from the MOXXI system. Drugs included in each category were defined by generic form and nationally standardized drug identification numbers and were updated to accommodate new medication in the therapeutic class during the study period.

Secondary outcome: adherence with prescribed treatment for uncomplicated hypertension

Adherence to drug treatment is adversely affected by higher out-of-pocket expenditures [16, 38,39,40,41, 55]. If the intervention produces meaningful reductions in the out-of-pocket expenditures for patients, then one would expect a corresponding improvement in treatment adherence. Treatment adherence was defined as proportion of days in which the patient had the expected supply of medication in the year following the visit in which the prescription was received [71, 72]. Records of prescribed, stopped, and dispensed drugs for each patient for the year following the first visit post-randomization and the 6 months prior to randomization were retrieved and used to create a drug by day matrix of expected and actual drug supply days as a function of prescribed therapy. When a patient was prescribed more than one drug in a therapeutic class, the mean adherence value was calculated for all drugs prescribed.

Patient characteristics

Since physicians rather than patients were randomized, systematic differences may exist between intervention and control patients that may influence the outcome, as the characteristics of the practice population differ among physicians. To adjust for these potential differences between physician practice populations, the following patient characteristics were collected: age, sex, estimated household income, drug insurance plan, current treatment costs (for prevalent patients), comorbidities, and health care use in the year before their first visit to the physicians after randomization. Patient age and sex were obtained from the RAMQ beneficiary data. Capacity to pay for prescriptions is an important determinant of physician’s choice of medication [29] as well as the likelihood of patient adherence to drug therapy [21,22,23, 33,34,35,36]. Capacity to pay was determined by estimating the average household income in the six-digit postal code area of the patient’s residence (approximately 36 households) based on Canadian census data [30]. Baseline anti-hypertensive treatment costs for prevalent patients were measured as the average annual costs of anti-hypertensive drugs dispensed in the 12 months prior to randomization. Comorbidity was assessed using the Charlson comorbidity index [61, 73, 74], based on diagnostic codes recorded in medical service claims and hospitalization discharge records in the 12 months prior to the patient’s first visit post-randomization. Baseline health care service use included the number of ER visits, number of hospitalizations, number of prescribed medications, and continuity of care with the primary care physician in the 12 months prior to the first visit post randomization for each patient using previously published algorithms [75].


Descriptive statistics were used to evaluate differences in the baseline characteristics of participating physicians and patients in the two arms of the trial. Study hypotheses were tested using an intention-to-treat analysis, whereby all consenting patients who made at least one visit to the study physician during the follow-up period were included in the analysis. For all patients, longitudinal information on health care utilization and mortality was available through linked provincial health administrative databases [76].

Logistic and generalized linear regression models were respectively used to analyze the effect of the intervention on the primary (diuretic prescribed) and secondary (treatment adherence) outcomes. To account for clustering of patients within physician, multivariate models were estimated within a generalized estimating equation (GEE) framework [77, 78], assuming an exchangeable correlation structure. Newly treated and currently treated hypertension patients were analyzed separately in relationship to the primary and secondary outcomes. The treatment effect (intervention vs. control) was represented as a binary level variable and was included in GEE models along with adjustment for baseline differences in patient demographics, comorbidity, household income, number of medications taken, and copayment plan, as well as physician-level differences in average costs of prescribed anti-hypertensives, years in practice, sex, practice volume, and propensity to prescribe diuretics. We suspected that both the patient-level differences as well as physician-level differences would affect physicians’ willingness to influence out-of-pocket costs [79,80,81]. In addition, to determine if the effect of the intervention was modified by these characteristics, we fit two-way interaction terms between each patient and physician characteristic with the intervention and assessed significance based on the model fit as well as the interaction term estimates.


Seventy-nine physicians were enrolled in this study. 25.1% of their 67,012 patients with hypertension were excluded on the first visit because they had complicated hypertension, were already using a diuretic, or were prescribed multiple anti-hypertensive therapies (Fig. 3). Of the remaining 50,163 patients, 60.8% either were not covered by the public drug plan or were receiving free medication. Of the remaining patients, 3592 had a diagnosis of uncomplicated hypertension and were prescribed anti-hypertensive treatment during the follow-up period by the study physicians. Seven physicians retired during the follow-up period and their data were included in the analysis.

Fig. 3
figure 3

Physicians and patients eligible and enrolled in the trial: February 2009 to July 2015

Study physicians in the intervention and control groups were similar in sex, language, number of years in practice, practice size, and daily patient visits, as well as the estimated household income of their patients (Table 1). The mean number of patients being treated for hypertension at the start of the study was 152.7 in the control and 187.6 in the intervention group. For these patients, the most commonly prescribed anti-hypertensive treatment was calcium channel blockers (control group 21.7%, intervention 19.1%) followed by diuretics and beta blockers. Mean annual costs of anti-hypertensive treatment were $370.90 in the control group and $385.70 in the intervention group. Annual out-of-pocket patient costs were similar.

Table 1 Characteristics of the 76 physicians in the intervention and control groups, as well as the baseline characteristics of all patients under their care

Most eligible patients with uncomplicated hypertension were between 60 to 75 years of age, female, in the middle-income bracket, and with few comorbidities that would increase the risk of mortality (Charlson index value of 0) (Table 2). Continuity of care was moderate, with 38.6% (intervention) and 41.1% (control) of all visits being made to the primary care physician, who was responsible for only 49.8% (control) to 50.0% (intervention) of all drugs prescribed. The proportion of patients with one or more ED visits and/or hospitalizations was similar in the intervention and control groups. Between 32.7% (intervention) and 37.9% (control) of patients were started on therapy in the follow-up (incident patients) period. For patients already prescribed anti-hypertensive medication (prevalent patients) at the start of the study, a mean of 1.3 different anti-hypertensives was prescribed concurrently; 94.3% (intervention) and 93.6% (control) of which were prescribed by the patient’s study physician.

Table 2 Baseline characteristics of the 3592 hypertension patients in the intervention and control groups between 2009 and 2015

Among newly treated patients, 26.6% were prescribed a diuretic in the intervention group compared to 19.8% in the control group, a significant increase of 65% (adjusted OR 1.65, p = 0.01) in the odds of prescribing diuretics in the intervention group after adjusting for patient and physician characteristics (Table 3). Among patients who had already been started on anti-hypertensive therapy before the start of the study, 15.5% (intervention) and 15.3% (control) were switched to diuretics, a non-statistically significant change in the odds of diuretic prescribing (OR 1.09, p = 0.60). Incident and prevalent patients in the intervention group were also more likely to be prescribed a single anti-hypertensive compared to the control for both newly treated (intervention 85.4%; control 84.0%) and currently treated patients (intervention 77.4%, control 74.5%), although this difference was not statistically significant. We assessed whether the intervention was modified by age, comorbidity, level of co-payment required by the drug insurance plan, estimated household income, and the patient’s co-payment plan. There was a significant modification of the effect of the intervention by patient age among the prevalent patients. Older patients who had already been started on anti-hypertensive therapy were less likely to have their anti-hypertensive medication switched to a diuretic than patients under the age of 65 (Fig. 4). At the physician level, years of experience, sex, practice volume, average costs of prescribed anti-hypertensives, and prior rates of diuretic prescribing were assessed. Only the number of years in practice modified the effect of the intervention for newly treated incident patients. Physicians in practice less than 15 years were much more likely to prescribe a diuretic (OR 10.69; 95% CI 1.49 to 76.64) than physicians in practice 15 to 25 years (OR 0.67; 95% CI 0.25 to 1.78) or in practice more than 25 years (OR 1.80; 95% CI 1.23 to 2.65).

Table 3 Outcome estimates of the 3592 study patients by usage status at the end of 60-month observations
Fig. 4
figure 4

Probability of being prescribed a diuretic among currently treated intervention and control patients as modified by the age of patients. Covariates adjusted include gender, age of patient, Charlson comorbidity index, household income, number of medications, co-payment amount (maximum vs. partial/minimum), cost of total anti-hypertensive medications prescribed at the physician’s practice, physician’s level of experience, and physician’s propensity to prescribe diuretics, as well as the interaction term between the age of patient and the intervention

In relationship to secondary outcomes, mean adherence in the year after the first visit was higher in the intervention group for both incident (intervention 57.7%; control 52.8%) and prevalent patients (intervention 72.1%; control 70.7%); however, the difference in adherence after adjusting for patient and physician characteristics was not statistically significant (Table 3). There was also no statistically significant difference in the overall annual costs or out-of-pocket costs for anti-hypertensive therapy in the intervention and control groups for either incident or prevalent patients.


It has been postulated that the provision of information to physicians about individual patient’s out-of-pocket medication costs would influence their prescribing decisions. This study was the first to test this hypothesis using a sophisticated e-prescribing system that incorporated information on a patient’s concurrent medication, their drug insurance plan coverage, adherence to current medications, and alternate evidence-based options to therapy. Our target was the treatment, both incident and prevalent, of uncomplicated hypertension in community practice. We showed a significant increase in the prescribing of diuretics, the lowest out-of-pocket option for uncomplicated hypertension, for newly diagnosed patients. Among patients who were already on treatment, there was a significant effect on switching younger but not older patients to diuretics from other anti-hypertensive agents. The effect of the intervention was significantly greater for younger physicians and those in practice less than 15 years. They were more likely to prescribe diuretics for newly treated patients compared to similar physicians in the control group. There was no significant effect of the intervention on treatment adherence, total annual anti-hypertensive drug costs, or out-of-pocket payments. Further inspection of medications dispensed to newly treated patients in the 12-month follow-up period revealed that, compared with the control group, a higher proportion of patients in the intervention group had medications from other classes of hypertensives added to their initial prescription of diuretics. This might explain the observed lack of an effect of the intervention on total annual anti-hypertensive drug costs or out-of-pocket payments despite its effect on diuretic prescribing.

We evaluated the impact of providing comparative out-of-pocket cost information to physicians on their selection of treatment using uncomplicated hypertension as a case exemplar. Treatment of hypertension may not have been the best model as there is rapidly evolving evidence and controversy about guidelines for hypertension management, particularly in treatment targets and first-line treatment choices for uncomplicated and complicated hypertension [70, 82]. Hypertension treatment guidelines have changed over time. Initially, diuretics or beta blockers were recommended as first-line treatment of uncomplicated hypertension, and subsequently, calcium channel blockers, angiotensin-converting enzyme inhibitors, or angiotensin 2 receptor blockers were added as first-line treatment [70]. Hypertension treatment guidelines are susceptible to stakeholder interests [83], and the level of evidence for many recommendations does not meet grade A criteria [84]. Of interest, the effect of the intervention was significantly greater for younger physicians who are more likely to adopt electronic health record systems and computerized decision-support recommendations [85, 86], whereas older physicians are more likely to rely on industry-sponsored events for continuing professional development [87]. An early qualitative evaluation of hypertension management decision-making in a subset of this cohort of physicians provided some interesting insights about the factors that influence physician treatment decisions in hypertension management. Physicians who were less likely to prescribe diuretics believed these drugs were less efficacious, had more side effects, preferred treatment that would rapidly achieve their therapeutic target, and relied more on their own experience than the scientific evidence [83]. For these physicians, educational interventions may have limited success in altering prescribing behavior, which may explain the high failure rate in interventions to improve the cost-effectiveness of anti-hypertensive treatment [44, 45, 88]. An alternate approach would be to provide patients with the same information about the costs and out-of-pocket payments for alternative equally effective therapy. In comparison to physician-targeted interventions, this approach has been very successful in reducing the use of benzodiazepines in older adults [89]. Patient-centered approaches, particularly in the case of individual financial burden, should be considered in future interventions to improve cost-effective therapy.

This study was to provide physicians with salient information about out-of-pocket costs and the alternate evidence-based lowest cost alternatives at the time of making treatment decisions (incident patients), or when opening up the chart for the visit (prevalent patients). For incident patients, the computerized decision support was interruptive; the screen concerning alternate treatment and differences in cost would open without physician action (Fig. 1). These forms of alerts have been shown to be more effective than non-interruptive alerts [45, 90]. The significant interaction between the age of the patient and the effect of the intervention among these prevalent patients suggests that the reluctance to switch current treatment for older patients may be due to the fact that they may have already failed a trial of diuretics or the patients declined the treatment change. Informally, physicians in the intervention arm indicated that they would have preferred to have drug cost and out-of-pocket information for more than one evidence-based option, allowing them to use their clinical judgment in determining what would be best for an individual patient. We might have modified the intervention to address this concern had we done more user testing prior to implementation. Other studies have shown that synthesizing information to enable physicians to more easily visualize the risks and benefits associated with different treatment options has been successfully used in altering the prescription of psychotropic drugs in older adults [91]. More extensive user testing and the provision of information on all out-of-pocket costs for alternate choices should be considered in future computerized decision-support interventions to reduce unnecessary drug costs.

There are several limitations that need to be considered. The physician cohort was a non-random sample of primary care physicians practicing in the province of Quebec. While this will not bias the internal validity of the study, the results may not be generalizable to other jurisdictions. Our eligibility criteria for enrolling patients was strict, limiting the diagnosis to uncomplicated hypertension, absence of other comorbidities that may alter treatment choices and receiving treatment only from the primary care physician involved in the study. Many patients who started on hypertensive treatment in the study physicians’ practices were excluded because the single criterion of a diagnosis of uncomplicated hypertension was missing. Since these diagnoses were derived from the codes submitted along with the medical services claims and these codes have been shown to have high specificity and low sensitivity, more patients may have been eligible for inclusion into the study. Physicians were randomized, not clinics in order to obtain a better balance in physician and patient characteristics between the intervention and control groups. While physicians who were co-located in the same clinic do not usually share patient management except for weekend call, it is plausible that an intervention physician may have discussed out-of-pocket costs with a control physician. If these discussions modified the behavior of control physicians, contamination will have reduced observed differences between the two groups.


In conclusion, the provision to physicians of comparative information on out-of-pocket payments for the treatment of uncomplicated hypertension resulted in a significant but modest impact on increasing the prescription of diuretics as initial therapy in newly diagnosed patients and the switch to diuretics from other agents in younger patients already on treatment. The impact of interventions to improve the cost-effectiveness of prescribing may be enhanced by also targeting patients with tools to participate in treatment decision-making and by providing physicians with comparative out-of-pocket information on all evidence-based alternatives.


  1. 1.

    Information CIfH. National health expenditure trends, 1985 to 2012. Ottawa: Canadian Institute for Health Information (CFHI); 2013.

  2. 2.

    PMPRB: Patented medicine prices review board annual report 2015. Edited by PMPRB. Ottawa: Patented Medicine Prices Review Board (PMPRB); 2016.

  3. 3.

    Institute Q. Outlook for global medicines through 2021. Parsippany: Quintiles IMS Institute; 2016.

  4. 4.

    OECD: Healthcare costs unsustainable in advanced economies without reform. 2015.

    Google Scholar 

  5. 5.

    Corporation AR, Mays J, Brenner M. Estimates of Medicare beneficiaries’ out-of-pocket drug spending in 2006. In: www.kf. org. Menlo Park: The Henry J.Kaiser Family Foundation; 2004.

    Google Scholar 

  6. 6.

    Huskamp HA, Keating NL. The new Medicare drug benefit: potential effects of pharmacy management tools on access to medications. In: The Henry J Kaiser Family Foundation. Menlo Park: The Henry J. Kaiser Family Foundation; 2004.

    Google Scholar 

  7. 7.

    Sarpatwari A, Avorn J, Kesselheim AS. State initiatives to control medication costs—can transparency legislation help? N Engl J Med. 2016;374(24):2301–4.

    PubMed  Article  Google Scholar 

  8. 8.

    Colla CH, Morden NE, Sequist TD, Schpero WL, Rosenthal MB. Choosing wisely: prevalence and correlates of low-value health care services in the United States. J Gen Intern Med. 2015;30(2):221–8.

    PubMed  Article  Google Scholar 

  9. 9.

    Berwick DM, Hackbarth AD. Eliminating waste in US health care. JAMA. 2012;307(14):1513–6.

    CAS  PubMed  Article  Google Scholar 

  10. 10.

    Newhouse JP. How much should Medicare pay for drugs? Health Aff (Millwood ). 2004;23(1):89–102.

    Article  Google Scholar 

  11. 11.

    Antos JR. Ensuring access to affordable drug coverage in Medicare. Health care financing review. 2005;27(2):103–12.

    PubMed  PubMed Central  Google Scholar 

  12. 12.

    Safran DG, Strollo MK, Guterman S, Li A, Rogers WH, Neuman P. Prescription coverage, use and spending before and after Part D implementation: a national longitudinal panel study. J Gen Intern Med. 2010;25(1):10–7.

    PubMed  Article  Google Scholar 

  13. 13.

    Fronstin P, Sepulveda MJ, Roebuck MC. Consumer-directed health plans reduce the long-term use of outpatient physician visits and prescription drugs. Health Aff (Project Hope). 2013;32(6):1126–34.

    Article  Google Scholar 

  14. 14.

    Goedken AM, Urmie JM, Farris KB, Doucette WR. Impact of cost sharing on prescription drugs used by Medicare beneficiaries. Res Social Adm Pharm. 2010;6(2):100–9.

    PubMed  Article  Google Scholar 

  15. 15.

    Gemmill MC, Thomson S, Mossialos E. What impact do prescription drug charges have on efficiency and equity? Evidence from high-income countries. Int J Equity Health. 2008;7:12.

    PubMed  PubMed Central  Article  Google Scholar 

  16. 16.

    Luiza VL, Chaves LA, Silva RM, Emmerick IC, Chaves GC, Fonseca de Araujo SC, Moraes EL, Oxman AD. Pharmaceutical policies: effects of cap and co-payment on rational use of medicines. Cochrane Database Syst Rev. 2015;5:Cd007017.

    Google Scholar 

  17. 17.

    Eaddy MT, Cook CL, O'Day K, Burch SP, Cantrell CR. How patient cost-sharing trends affect adherence and outcomes: a literature review. Pharm Ther. 2012;37(1):45–55.

    Google Scholar 

  18. 18.

    Daniel H. Stemming the escalating cost of prescription drugs: a position paper of the American College of Physicians. Ann Intern Med. 2016;165:50–2.

    Article  Google Scholar 

  19. 19.

    Morgan S, Hanley G, Greyson D. Comparison of tiered formularies and reference pricing policies: a systematic review. Open Med. 2009;3(3):e131–9.

    PubMed  PubMed Central  Google Scholar 

  20. 20.

    Acosta A, Ciapponi A, Aaserud M, Vietto V, Austvoll-Dahlgren A, Kosters JP, Vacca C, Machado M, Diaz Ayala DH, Oxman AD. Pharmaceutical policies: effects of reference pricing, other pricing, and purchasing policies. Cochrane Database Syst Rev. 2014;10:Cd005979.

    Google Scholar 

  21. 21.

    Kamal-Bahl S, Briesacher B. How do incentive-based formularies influence drug selection and spending for hypertension? Health Aff. 2004;23(1):227–36.

    Article  Google Scholar 

  22. 22.

    Gleason PP, Gunderson BW, Gericke KR. Are incentive-based formularies inversely associated with drug utilization in managed care? Ann Pharmacother. 2005;39(2):339–45.

    PubMed  Article  Google Scholar 

  23. 23.

    Briesacher B, Kamal-Bahl S, Hochberg M, Orwig D, Kahler KH. Three-tiered-copayment drug coverage and use of nonsteroidal anti-inflammatory drugs. Arch Intern Med. 2004;164(15):1679–84.

    PubMed  Article  Google Scholar 

  24. 24.

    Allan GM, Lexchin J, Wiebe N. Physician awareness of drug cost: a systematic review. PLoS Med. 2007;4(9):e283.

    PubMed  PubMed Central  Article  Google Scholar 

  25. 25.

    Shrank WH, Young HN, Ettner SL, Glassman P, Asch SM, Kravitz RL. Do the incentives in 3-tier pharmaceutical benefit plans operate as intended? Results from a physician leadership survey. Am J Manag Care. 2005;11(1):16–22.

    PubMed  Google Scholar 

  26. 26.

    Allan GM, Innes GD. Do family physicians know the costs of medical care? Survey in British Columbia. Can Fam Physician. 2004;50:263–70.

    PubMed  PubMed Central  Google Scholar 

  27. 27.

    Watkins C, Harvey I, Carthy P, Moore L, Robinson E, Brawn R. Attitudes and behaviour of general practitioners and their prescribing costs: a national cross sectional survey. Qual Saf Health Care. 2003;12(1):29–34.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  28. 28.

    Reichert S, Simon T, Halm EA. Physicians’ attitudes about prescribing and knowledge of the costs of common medications. Am Med Assoc. 2000;160:2799–803.

    CAS  Google Scholar 

  29. 29.

    Alexander GC, Casalino LP, Meltzer DO. Physician strategies to reduce patients’ out-of-pocket prescription costs. Arch Intern Med. 2005;165(6):633–6.

    PubMed  Article  Google Scholar 

  30. 30.

    Tamblyn R, Laprise R, Hanley JA, Abrahamowicz M, Scott S, Mayo N, Hurley J, Grad R, Latimer E, Perreault R, et al. Adverse events associated with prescription drug cost-sharing among poor and elderly persons. JAMA. 2001;285(4):421–9.

    CAS  PubMed  Article  Google Scholar 

  31. 31.

    Tamblyn R. The impact of pharmacotherapy policy: a case study. Can J Clin Pharmacol. 2001;8 Suppl A:39A–44A.

    CAS  PubMed  Google Scholar 

  32. 32.

    Schneeweiss S, Walker AM, Glynn RJ, Maclure M, Dormuth C, Soumerai SB. Outcomes of reference pricing for angiotensin-converting-enzyme inhibitors. N Engl J Med. 2002;346(11):822–9.

    PubMed  Article  Google Scholar 

  33. 33.

    Huskamp HA, Deverka PA, Epstein AM, Epstein RS, McGuigan KA, Frank RG. The effect of incentive-based formularies on prescription-drug utilization and spending. N Engl J Med. 2003;349(23):2224–32.

    CAS  PubMed  Article  Google Scholar 

  34. 34.

    Rector TS, Finch MD, Danzon PM, Pauly MV, Manda BS. Effect of tiered prescription copayments on the use of preferred brand medications. Med Care. 2003;41(3):398–406.

    PubMed  Google Scholar 

  35. 35.

    Joyce GF, Escarce JJ, Solomon MD, Goldman DP. Employer drug benefit plans and spending on prescription drugs. JAMA. 2002;288(14):1733–9.

    PubMed  Article  Google Scholar 

  36. 36.

    Safran DG, Neuman P, Schoen C, Kitchman MS, Wilson IB, Cooper B, Li A, Chang H, Rogers WH. Prescription drug coverage and seniors: findings from a 2003 national survey. Health Aff (Millwood). 2005;W5–166. doi:10.1377/hlthaff.w5.152.

  37. 37.

    Tseng CW, Brook RH, Keeler E, Steers WN, Mangione CM. Cost-lowering strategies used by Medicare beneficiaries who exceed drug benefit caps and have a gap in drug coverage. JAMA. 2004;292(8):952–60.

    CAS  PubMed  Article  Google Scholar 

  38. 38.

    Heisler M, Langa KM, Eby EL, Fendrick AM, Kabeto MU, Piette JD. The health effects of restricting prescription medication use because of cost. Med Care. 2004;42(7):626–34.

    PubMed  Article  Google Scholar 

  39. 39.

    Lexchin J, Grootendorst P. Effects of prescription drug user fees on drug and health services use and on health status in vulnerable populations: a systematic review of the evidence. Int J Health Serv. 2004;34(1):101–22.

    PubMed  Article  Google Scholar 

  40. 40.

    Mojtabai R, Olfson M. Medication costs, adherence, and health outcomes among Medicare beneficiaries. Health Aff (Millwood ). 2003;22(4):220–9.

    Article  Google Scholar 

  41. 41.

    Doshi JA, Brandt N, Stuart B. The impact of drug coverage on COX-2 inhibitor use in Medicare. Health Aff. 2004;W4–105.

  42. 42.

    Dragomir A, Cote R, Roy L, Blais L, Lalonde L, Berard A, Perreault S. Impact of adherence to antihypertensive agents on clinical outcomes and hospitalization costs. Med Care. 2010;48(5):418–25.

    PubMed  Article  Google Scholar 

  43. 43.

    Despres F, Forget A, Kettani FZ, Blais L. Impact of patient reimbursement timing and patient out-of-pocket expenses on medication adherence in patients covered by private drug insurance plans. J Manag Care Spec Pharm. 2016;22(5):539–47.

    PubMed  Article  Google Scholar 

  44. 44.

    Tseng CW, Lin GA, Davis J, Taira DA, Yazdany J, He Q, Chen R, Imamura A, Dudley RA. Giving formulary and drug cost information to providers and impact on medication cost and use: a longitudinal non-randomized study. BMC Health Serv Res. 2016;16(1):499.

    PubMed  PubMed Central  Article  Google Scholar 

  45. 45.

    Pevnick JM, Li N, Asch SM, Jackevicius CA, Bell DS. Effect of electronic prescribing with formulary decision support on medication tier, copayments, and adherence. BMC Med Inform Decis Mak. 2014;14:79.

    PubMed  PubMed Central  Article  Google Scholar 

  46. 46.

    McMullin ST, Lonergan TP, Rynearson CS, Doerr TD, Veregge PA, Scanlan ES. Impact of an evidence-based computerized decision support system on primary care prescription costs. Ann Fam Med. 2004;2(5):494–8.

    PubMed  PubMed Central  Article  Google Scholar 

  47. 47.

    McMullin ST, Lonergan TP, Rynearson CS. Twelve-month drug cost savings related to use of an electronic prescribing system with integrated decision support in primary care. J Manag Care Pharm. 2005;11(4):322–32.

    PubMed  Google Scholar 

  48. 48.

    Ross SM, Papshev D, Murphy EL, Sternberg DJ, Taylor J, Barg R. Effects of electronic prescribing on formulary compliance and generic drug utilization in the ambulatory care setting: a retrospective analysis of administrative claims data. J Manag Care Pharm. 2005;11(5):410–5.

    PubMed  Google Scholar 

  49. 49.

    Wilson IB, Rogers WH, Chang H, Safran DG. Cost-related skipping of medications and other treatments among Medicare beneficiaries between 1998 and 2000. Results of a national study. J Gen Intern Med. 2005;20(8):715–20.

    PubMed  PubMed Central  Article  Google Scholar 

  50. 50.

    Lundin D. Moral hazard in physician prescription behavior. J Health Econ. 2000;19(5):639–62.

    CAS  PubMed  Article  Google Scholar 

  51. 51.

    Wing LM, Reid CM, Ryan P, Beilin LJ, Brown MA, Jennings GL, Johnston CI, McNeil JJ, Macdonald GJ, Marley JE, et al. A comparison of outcomes with angiotensin-converting–enzyme inhibitors and diuretics for hypertension in the elderly. N Engl J Med. 2003;348(7):583–92.

    CAS  PubMed  Article  Google Scholar 

  52. 52.

    Group. AOaCftACR. Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic: the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). JAMA. 2002;288(23):2981–97.

    Article  Google Scholar 

  53. 53.

    Psaty BM, Lumley T, Furberg CD, Schellenbaum G, Pahor M, Alderman MH, Weiss NS. Health outcomes associated with various antihypertensive therapies used as first-line agents: a network meta-analysis. JAMA. 2003;289(19):2534–44.

    CAS  PubMed  Article  Google Scholar 

  54. 54.

    Fischer MA, Avorn J. Economic implications of evidence-based prescribing for hypertension: can better care cost less? JAMA. 2004;291(15):1850–6.

    CAS  PubMed  Article  Google Scholar 

  55. 55.

    Morgan S, McMahon M, Lam J, Mooney D, Raymond C. The Canadian Rx Atlas. Vancouver; 2005.

  56. 56.

    Morgan S, Bassett KL, Wright JM, Yan L. First-line first? Trends in thiazide prescribing for hypertensive seniors. PLoS Med. 2005;2(4):e80.

    PubMed  PubMed Central  Article  Google Scholar 

  57. 57.

    Wilchesky M, Tamblyn RM, Huang A. Validation of diagnostic codes in medical services claims data. Can J Clin Pharmacol. 2001;8(1):39.

    Google Scholar 

  58. 58.

    Tamblyn R, Reid T, Mayo N, McLeod P, Churchill-Smith M. Using medical services claims to assess injuries in the elderly: the sensitivity of diagnostic and procedure codes for injury ascertainment. J Clin Epidemiol. 2000;53(2):183–94.

    CAS  PubMed  Article  Google Scholar 

  59. 59.

    Levy AR, Tamblyn R, Fitchett D, McLeod P, Hanley J. Coding accuracy of hospital discharge data for elderly survivors of myocardial infarction. Can J Cardiol. 1999;15(11):1277–82.

    CAS  PubMed  Google Scholar 

  60. 60.

    Tamblyn RM, Lavoie G, Petrella L, Monette J. The use of prescription claims databases in Pharmacoepidemiological research: the accuracy and comprehensiveness of the prescription claims database in Quebec. J Clin Epidemiol. 1995;48(8):999–1009.

    CAS  PubMed  Article  Google Scholar 

  61. 61.

    Wilchesky M, Tamblyn RM, Huang A. Validation of diagnostic codes within medical services claims. J Clin Epidemiol. 2004;57(2):131–41.

    PubMed  Article  Google Scholar 

  62. 62.

    Tamblyn R. The new millennium model for health care and research. In: Downey J, Claxton L, editors. Innovation: essays by leading Canadian researchers. Toronto: Key Porter Books Ltd.; 2003. p. 184–93.

    Google Scholar 

  63. 63.

    Tamblyn R, Huang A, Kawasumi Y, Bartlett-Esquilant G, Grad R, Jacques A, Dawes M, Abrahamowicz M, Perreault R, Taylor L, et al. The development and evaluation of an integrated electronic prescribing and drug management system for primary care. J Am Med Inform Assoc. 2006;13(2):148–59.

    PubMed  PubMed Central  Article  Google Scholar 

  64. 64.

    Apkon M, Singhaviranon P. Impact of an electronic information system on physician workflow and data collection in the intensive care unit. Intensive Care Med. 2001;27(1):122–30.

    CAS  PubMed  Article  Google Scholar 

  65. 65.

    Teich JM, Spurr CD, Schmiz JL, O’Connell EM, Thomas D. Enhancement of clinician workflow with computer order entry. Proc Annu Symp Comput Appl Med Care. 1995:459–63.

  66. 66.

    Webster C, Copenhaver J. Structured data entry in a workflow-enabled electronic patient record. J Med Pract Manage. 2001;17(3):157–61.

    CAS  PubMed  Google Scholar 

  67. 67.

    Tamblyn R, Huang A, Taylor L, Kawasumi Y, Bartlett G, Grad R, Jacques A, Dawes M, Abrahamowicz M, Perreault R, et al. A randomized trial of the effectiveness of on-demand versus computer-triggered drug decision support in primary care. J Am Med Inform Assoc. 2008;15(4):430–8.

    PubMed  PubMed Central  Article  Google Scholar 

  68. 68.

    Ordre des pharmaciens du Québec: Foire aux questions - Pratique professionelle []. Accessed 31 Oct 2017.

  69. 69.

    Kawamoto K, Houlihan CA, Balas EA, Lobach DF. Improving clinical practice using clinical decision support systems: a systematic review of trials to identify features critical to success. BMJ. 2005;330(7494):765.

    PubMed  PubMed Central  Article  Google Scholar 

  70. 70.

    Leung AA, Nerenberg K, Daskalopoulou SS, McBrien K, Zarnke KB, Dasgupta K, Cloutier L, Gelfer M, Lamarre-Cliche M, Milot A, et al. Hypertension Canada’s 2016 Canadian hypertension education program guidelines for blood pressure measurement, diagnosis, assessment of risk, prevention, and treatment of hypertension. Can J Cardiol. 2016;32(5):569–88.

    PubMed  Article  Google Scholar 

  71. 71.

    Tamblyn R, Eguale T, Huang A, Winslade N, Doran P. The incidence and determinants of primary nonadherence with prescribed medication in primary care: a cohort study. Ann Intern Med. 2014;160(7):441–50.

    PubMed  Article  Google Scholar 

  72. 72.

    Trabulsi N, Riedel K, Winslade N, Gregoire JP, Meterissian S, Abrahamovicz M, Tamblyn R, Mayo N, Meguerditchian A. Adherence to anti-estrogen therapy in seniors with breast cancer: how well are we doing? Breast J. 2014;20(6):632–8.

    CAS  PubMed  Article  Google Scholar 

  73. 73.

    Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373–83.

    CAS  PubMed  Article  Google Scholar 

  74. 74.

    Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45(6):613–9.

    CAS  PubMed  Article  Google Scholar 

  75. 75.

    Tamblyn R, Abrahamowicz M, Brailovsky C, Grand’Maison P, Lescop J, Norcini J, Girard N, Haggerty J. The association between licensing examination scores and resource use and quality of care in primary care practice. JAMA. 1998;280(11):989–96.

    CAS  PubMed  Article  Google Scholar 

  76. 76.

    Murphy SN, Gainer V, Chueh HC. A visual interface designed for novice users to find research patient cohorts in a large biomedical database. AMIA Annu Symp Proc. 2003;2003:489–93.

  77. 77.

    Donner A, Klar N. Design and analysis of cluster randomization trials in health research. London: Arnold Publishers; 2000.

    Google Scholar 

  78. 78.

    Chuang J-H, Hripcsak G, Jenders RA. Considering clustering: a methodological review of clinical decision support system studies. In: 2000 2000; 2000. p. 146–50.

    Google Scholar 

  79. 79.

    Davidson W, Malloy W, Somers G, Bédard M. Relationships between physician practice characteristics and prescribing behaviour for the elderly in New Brunswick. Can Med Assoc J. 1994;150:917.

    CAS  Google Scholar 

  80. 80.

    Tamblyn R, McLeod P, Hanley JA, Girard N, Hurley J. Physician and practice characteristics associated with the early utilization of new prescription drugs. Med Care. 2003;41(8):895–908.

    PubMed  Article  Google Scholar 

  81. 81.

    Lubloy A. Factors affecting the uptake of new medicines: a systematic literature review. BMC Health Serv Res. 2014;14:469.

    PubMed  PubMed Central  Article  Google Scholar 

  82. 82.

    Dasgupta K, Quinn RR, Zarnke KB, Rabi DM, Ravani P, Daskalopoulou SS, Rabkin SW, Trudeau L, Feldman RD, Cloutier L, et al. The 2014 Canadian hypertension education program recommendations for blood pressure measurement, diagnosis, assessment of risk, prevention, and treatment of hypertension. Can J Cardiol. 2014;30(5):485–501.

    PubMed  Article  Google Scholar 

  83. 83.

    Rochefort CM, Morlec J, Tamblyn RM. What differentiates primary care physicians who predominantly prescribe diuretics for treating mild to moderate hypertension from those who do not? A comparative qualitative study. BMC Fam Pract. 2012;13:9.

    PubMed  PubMed Central  Article  Google Scholar 

  84. 84.

    James PA, Oparil S, Carter BL, Cushman WC, Dennison-Himmelfarb C, Handler J, Lackland DT, LeFevre ML, MacKenzie TD, Ogedegbe O, et al. 2014 evidence-based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA. 2014;311(5):507–20.

    CAS  PubMed  Article  Google Scholar 

  85. 85.

    Jamoom E, Beatty P, Bercovitz A, Woodwell D, Palso K, Rechtsteiner E. Physician adoption of electronic health record systems: United States. NCHS data brief. 2011;2012(98):1–8.

    Google Scholar 

  86. 86.

    Menachemi N, Powers TL, Brooks RG. Physician and practice characteristics associated with longitudinal increases in electronic health records adoption. J Healthc Manag. 2011;56(3):183–97. discussion 197-188

    PubMed  Article  Google Scholar 

  87. 87.

    Wazana A. Physicians and the pharmaceutical industry: is a gift ever just a gift? JAMA. 2000;283(3):373–80.

    CAS  PubMed  Article  Google Scholar 

  88. 88.

    Zwarenstein M, Grimshaw JM, Presseau J, Francis JJ, Godin G, Johnston M, Eccles MP, Tetroe J, Shiller SK, Croxford R, et al. Printed educational messages fail to increase use of thiazides as first-line medication for hypertension in primary care: a cluster randomized controlled trial [ISRCTN72772651]. Implement Sci. 2016;11(1):124.

    PubMed  PubMed Central  Article  Google Scholar 

  89. 89.

    Tannenbaum C, Martin P, Tamblyn R, Benedetti A, Ahmed S. Reduction of inappropriate benzodiazepine prescriptions among older adults through direct patient education the EMPOWER cluster randomized trial. JAMA Intern Med. 2014;174(6):890–8.

    PubMed  Article  Google Scholar 

  90. 90.

    Pevnick JM, Li X, Grein J, Bell DS, Silka P. A retrospective analysis of interruptive versus non-interruptive clinical decision support for identification of patients needing contact isolation. Appl Clin Inform. 2013;4(4):569–82.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  91. 91.

    Tamblyn R, Eguale T, Buckeridge DL, Huang A, Hanley J, Reidel K, Shi S, Winslade N. The effectiveness of a new generation of computerized drug alerts in reducing the risk of injury from drug side effects: a cluster randomized trial. J Am Med Inform Assoc. 2012;19(4):635–43.

    PubMed  PubMed Central  Article  Google Scholar 

Download references


Not applicable


This study was funded by the Canadian Institutes of Health Research (Award # 82331).

Availability of data and materials

Data for this study is not available for public access due to the sensitivity of the data.

Author information




RT, NW, and AH participated in the conception and execution of the study. TM conducted the analysis. RT, NM, AH, CQ, and TM all participated in the preparation and final approval of the manuscript.

Corresponding author

Correspondence to Robyn Tamblyn.

Ethics declarations

Ethics approval and consent to participate

Ethics approval was obtained and maintained throughout the trial and study from the McGill Institutional Review Board (#A07-B27-06B). All patients included in the study provided consent to be included in the trial.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Tamblyn, R., Winslade, N., Qian, C.J. et al. What is in your wallet? A cluster randomized trial of the effects of showing comparative patient out-of-pocket costs on primary care prescribing for uncomplicated hypertension. Implementation Sci 13, 7 (2018).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI:


  • Computerized decision support
  • Electronic prescribing
  • Cost-effectiveness
  • Anti-hypertensive agents