Table 1 Characteristics of included studies

Adams, Soumerai, Ross-Degnan, 2001 [15]

 - Narrative review (critical review)

 - US Medicare population (65+ years)

 - Annual Review of Public Health

The effect of drug coverage on drug utilization, health outcomes, and health care costs in the Medicare population.

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: no

- grey literature: yes

- year of last search: not reported

- # of studies included: total, 37; drugs/cost-sharing/ins, 37; Canada, 0

- duplicate study selection and data extraction: unclear

No formal quality assessment conducted. Criteria were used to assess the validity of findings including study design, appropriateness of study population, data quality and availability, reliability of measures of association and adequacy of statistical analysis. The strengths and limitations of study designs were generally described.

- no ‘a priori’ design;

- non-systematic search strategy;

- no/unclear duplicate study selection and data extraction;

- list of excluded studies not provided;

- unclear screening and data extraction process;

- study characteristics of studies not provided;

- no formal quality assessment of included studies.

Harten, Ballantyne, 2004 [16]

 - Narrative review (review)

 - General population (Canadians)

 - Journal of Pharmaceutical Finance, Economics, and Policy

Canadian evidence of the effects of cost-sharing mechanisms of provincial drug benefit programs on program expenditures, drug utilization and patient health.

 - drug use: yes

 - healthcare use: no

 - health: yes

- a priori’ design: no

- search comprehensive: no

- grey literature: yes

- year of last search: 2002

- # of studies included: total, 7; drugs/cost-sharing/ins, 7; Canada, 7

- duplicate study selection and data extraction: unclear

No formal quality assessment conducted. Limitations of included studies generally discussed.

- no ‘a priori’ design;

- unclear duplicate study selection and data extraction;

- non-systematic search strategy;

- list of excluded studies not provided;

- no formal quality assessment of included studies.

Lexchin, Grootendorst, 2004 [17]

 - Narrative review (systematic review)

 - The poor and chronically ill

 - International Journal of Health Services

The effect of drug user fees on drug use and related outcomes in vulnerable populations (the poor and chronically ill).

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: yes

- grey literature: yes

- year of last search: 2002

- # of studies included: total, 24; drugs/cost-sharing/ins, 24; Canada, 5

- duplicate study selection and data extraction: yes

None

- no ‘a priori’ design;

- no quality assessment of included studies.

Rice, Matsuoka, 2004 [18]

 - Narrative review (review)

 - Seniors

 - Medical Care Research & Review

Impact of cost-sharing for medical services and prescription drugs on service use and health status of seniors.

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: no

- grey literature: no

- year of last search: not reported

- # of studies included: total, 22; drugs/cost-sharing/ins, 16; Canada, 4

- duplicate study selection and data extraction: unclear

No formal quality assessment conducted. The limitations of included studies were generally discussed.

- no ‘a priori’ design;

- grey literature not searched;

- unclear screening and data extraction process (inclusion and exclusion criteria not stated);

- list of excluded studies not provided;

- no formal quality assessment of included studies;

- narrow inclusion criteria (only studies from US and Canada were included).

Gibson, Ozminkowsky, Goetzel, 2005 [19]

 - Narrative review (review)

 - General population

 - American Journal of Managed Care

Do patients respond to increased cost-sharing by substituting less expensive alternatives for medications with higher levels of copayments or coinsurance?

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: no

- grey literature: no

- year of last search: 2005

- # of studies included: total, 30; drugs/cost-sharing/ins, 30; Canada, 4

- duplicate study selection and data extraction: unclear

None

- no ‘a priori’ design;

- narrow inclusion criteria (only studies from US and Canada and that used claims-based data sources were included);

- unclear inclusion/exclusion criteria;

- list of excluded studies not provided;

- study characteristics not clearly presented and/or synthesized;

- no formal quality assessment of included studies.

Maio, Pizzi, Roumm, 2005 [20]

 - Narrative review (review)

 - Seniors

 - The Milbank Quarterly

Among seniors, the effects of cost-sharing mechanisms and administrative mechanisms on prescription drug utilization and/or expenditures, other health services, underuse of effective medications, clinical outcomes, adverse events; and on the subject’s behaviour, such as voluntary disenrollment.

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: unclear

- grey literature: no

- year of last search: 2003

- # of studies included: total, 16; drugs/cost-sharing/ins, 7; Canada, 3

- duplicate study selection and data extraction: unclear

No formal quality assessment. The quality of quasi-experimental studies was generally discussed but not of randomized studies.

- no ‘a priori’ design;

- no/unclear duplicate study selection and data extraction;

- non-systematic search strategy;

- list of excluded studies not provided;

- grey literature not searched;

- no formal quality assessment of included studies.

Briesacher, Gurwitz, Soumerai, 2007 [21]

 - Narrative review (review)

 - General population

 - J Gen Intern Med

To identify patient-, medication-, and provider-level factors that influence the relationship between medication adherence and medication costs.

 - drug use: yes

 - healthcare use: no

 - health: no

- a priori’ design: no

- search comprehensive: yes

- grey literature: no

- year of last search: 2006

- # of studies included: total, 19; drugs/cost-sharing/ins, 17; Canada, 0

- duplicate study selection and data extraction: no

No formal quality assessment of included studies. Limitations of included studies were generally discussed.

- no ‘a priori’ design;

- no/unclear duplicate study selection and data extraction;

- grey literature not searched;

- study characteristics of included studies not provided;

- no formal quality assessment of included studies.

Gemmil, Costa-Font, McGuire, 2007 [22]

 - Meta-regression (meta-regression)

 - General population

 - Health Economics

To determine an estimate for drug-price elasticity using meta-regression analysis.

 - drug use: yes

 - healthcare use: no

 - health: no

- a priori’ design: no

- search comprehensive: no

- grey literature: no

- year of last search: not reported

- # of studies included: total, 31; drugs/cost-sharing/ins, 31; Canada, 6

- duplicate study selection and data extraction: unclear

No formal quality assessment of included studies. Limitations of included studies were generally discussed.

- no ‘a priori’ design;

- no/unclear duplicate study selection and data extraction;

- search strategy poorly described;

- grey literature not searched;

- study characteristics of included studies not provided;

- poorly justified or unclear exclusion criteria;

- no formal quality assessment of included studies.

Goldman, Joyce, Zheng, 2007 [23]

 - Narrative review (review)

 - General population

 - JAMA

Associations among cost-sharing features of prescription drug benefits and use of prescription drugs, use of non-pharmaceutical services, and health outcomes.

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: no

- grey literature: no

- year of last search: not reported

- # of studies included: total, 132; drugs/cost-sharing/ins, 132; Canada, 26

- duplicate study selection and data extraction: unclear

No formal quality assessment of included studies. Limitations of included studies were generally discussed.

- no ‘a priori’ design;

- non-systematic search strategy;

- grey literature not searched;

- no formal quality assessment of included studies.

- no/unclear duplicate study selection and data extraction;

- list of excluded studies not provided.

Gemmil, Thomson, Mossialos, 2008 [24]

 - Narrative review (review)

 - General population

 - International Journal for Equity in Health

The impact of prescription drug charges on efficiency and equity.

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: yes

- grey literature: yes

- year of last search: 2006

- # of studies included: total, 173; drugs/cost-sharing/ins, 173; Canada, 28

- duplicate study selection and data extraction: unclear

None; ‘quality’ was assessed by looking at study design, type of data analyzed, and techniques used for analysis but quality was not assessed beyond that.

- no ‘a priori’ design;

- no/unclear duplicate study selection and data extraction;

- study characteristics of included studies not provided.

- list of excluded studies not provided;

- no formal quality assessment of included studies.

Remler, Greene, 2009 [25]

 - Narrative review (review)

 - General population

 - Annual Review of Public Health

To determine the effects of cap and co-payment policies on rational use of medicines, healthcare utilization, health outcomes, and costs.

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: unclear

- grey literature: yes

- year of last search: not reported

- # of studies included: unclear

- duplicate study selection and data extraction: unclear

None

- no ‘a priori’ design;

- no/unclear duplicate study selection and data extraction;

- non-systematic search strategy; search strategy poorly described;

- unclear screening and data extraction process;

- list of included and excluded studies not provided;

- study characteristics of included studies not provided;

- no formal quality assessment of included studies.

Green, Maclure, et al., 2010 [26]

 - Narrative review (systematic)

 - General population

 - Report: The Cochrane Library

The effects of a pharmaceutical policy restricting the reimbursement of selected medications on drug use, health care utilization, health outcomes, and costs.

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: yes

- search comprehensive: yes

- grey literature: yes

- year of last search: 2009

- # of studies included: total, 29; drugs/cost-sharing/ins, 29; Canada, 11

- duplicate study selection and data extraction: yes

Included studies were appraised using Cochrane EPOC criteria for interrupted time series.(1)

- restrictive inclusion criteria limits the usefulness of the review.

Holst, 2010 [27]

 - Narrative review (systematic in-depth review)

 - General population

 - Working paper: Wissenschaftszentrum Berlin für Sozialforschung (WZB)

Does direct patient cost-sharing improve the efficiency of use of resources in health care? What effects does it have on social inequality of health opportunities in the population, and the political goal of reducing this?

 - drug use: yes

 - healthcare use: yes

 - health: no

- a priori’ design: no

- search comprehensive: unclear

- grey literature: yes

- year of last search: not reported

- # of studies included: not clearly reported

- duplicate study selection and data extraction: unclear

None

- no ‘a priori’ design;

- search strategy poorly described;

- unclear screening and data extraction process (inclusion and exclusion criteria not stated);

- list of included and excluded studies not provided;

- study characteristics of included studies not provided;

- no formal quality assessment of included studies.

Polinski, Kilabuk, et al., 2010 [28]

 - Narrative review (systematic)

 - US Medicare population (65+ years)

 - Journal of the American Geriatrics Society

To assess the extent to which Medicare Part D’s cost-sharing provisions and drug coverage rules affected the under- and overuse of specific drugs and classes.

 - drug use: yes

 - healthcare use: no

 - health: no

- a priori’ design: no

- search comprehensive: no

- grey literature: no

- year of last search: 2009

- # of studies included: total, 26; drugs/cost-sharing/ins, 26; Canada, 0

- duplicate study selection and data extraction: yes

The Newcastle-Ottawa Scale for cohort studies was used.(2) Unclear how domains were operationalized and assessed.

- no ‘a priori’ design;

- non-systematic search strategy;

- grey literature not searched;

- unclear exclusion criteria;

- list of excluded studies not provided;

- formal quality assessment of included studies poorly described and discussed; only global ratings provided; unclear how any of the domains were operationalized and assessed.

Swartz, 2010 [29]

 - Narrative review (synthesis)

 - General population

 - The Synthesis Project (The Robert Wood Johnson Foundation)

What is known and unknown about the effects of consumer cost sharing?

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: no

- grey literature: unclear

- year of last search: unclear

- # of studies included: total, unclear; drugs/cost-sharing/ins, unclear; Canada, unclear

- duplicate study selection and data extraction: no

No formal quality assessment conducted. However, in general, greater weight was given to studies that used data from natural experiments with credible comparison groups, as well as studies that relied on data from larger numbers of people and from people who were representative of subgroups of people.

- no ‘a priori’ design;

- list of excluded studies not provided;

- no/unclear duplicate study selection and data extraction;

- search strategy poorly described;

- poorly justified or unclear exclusion criteria;

- study characteristics of included studies not provided;

- no formal quality assessment of included studies;

Baicker, Goldman, 2011 [30]

 - Narrative review (review)

 - General population

 - Journal of Economic Perspectives

To determine the relationship between patient cost-sharing and healthcare spending growth.

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: unclear

- grey literature: not reported

- year of last search: not reported

- # of studies included: total, not reported; drugs/cost-sharing/ins, not reported; Canada, 1

- duplicate study selection and data extraction: unclear

None

- no ‘a priori’ design;

- no/unclear duplicate study selection and data extraction;

- search strategy not described;

- list of included and excluded studies not provided;

- no quality assessment of included studies;

- study characteristics of included studies not provided;

- focus on US studies limits the usefulness of the review.

Polinski, Donohue, et al., 2011 [31]

 - Narrative review (systematic)

 - US Medicare population (65+ years)

 - Journal of the American Geriatrics Society

The extent to which Medicare Part D’s cost-sharing provisions and drug coverage rules affected the under- and overuse of specific drugs and classes.

 - drug use: yes

 - healthcare use: no

 - health: no

- a priori’ design: no

- search comprehensive: no

- grey literature: no

- year of last search: 2010

- # of studies included: total, 19; drugs/cost-sharing/ins, 19; Canada, 0

- duplicate study selection and data extraction: yes

The Newcastle-Ottawa Scale for cohort studies was used.(2) Unclear how domains were operationalized and assessed.

- no ‘a priori’ design;

- unclear inclusion/exclusion criteria;

- non-systematic search strategy;

- search strategy poorly described;

- grey literature not searched;

- formal quality assessment of included studies poorly described and discussed; only global ratings provided; unclear how any of the domains were operationalized and assessed.

Eaddy, Cook, et al., 2012 [32]

 - Narrative review (review)

 - General population

 - Pharmacy and Therapeutics

To assess the relationship between patient cost-sharing, medication adherence, clinical, utilization, and economic outcomes.

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: yes

- grey literature: no

- year of last search: 2008

- # of studies included: total, 160; drugs/cost-sharing/ins, 160; Canada, 17

- duplicate study selection and data extraction: unclear

No formal quality assessment of included studies. Limitations of included studies were generally discussed.

- no ‘a priori’ design;

- unclear duplicate study selection and data extraction;

- search strategy poorly described;

- grey literature not searched;

- unclear screening and data extraction process (inclusion and exclusion criteria not clearly stated);

- study characteristics of included studies not provided;

- list of excluded studies not provided;

- no formal quality assessment of included studies.

Lemstra, Blackburn et al., 2012 [33]

 - Meta-analysis (meta-analysis)

 - Statin users

 - Canadian Journal of Cardiology

To provide estimates of risk indicators associated with nonadherence to statin medications.

 - drug use: yes

 - healthcare use: no

 - health: no

- a priori’ design: no

- search comprehensive: yes

- grey literature: no

- year of last search: 2011

- # of studies included: total, 67; drugs/cost-sharing/ins, 6; Canada, 1

- duplicate study selection and data extraction: unclear

RCTs: Used the Delphi list to evaluate quality;

Observational studies: assessed patient selection process, criteria for inclusion and exclusion, patient identification process, comparative information for the patients who were not enrolled in the study, attrition, lost-to follow-up, statistical analysis, and controlling for confounding variables.

A score of 5/9 and 5/8 was required for an RCT and an observational study, respectively, to be included.

Unclear how domains were operationalized and assessed.

- no ‘a priori’ design;

- no/unclear duplicate study selection and data extraction;

- grey literature not searched;

- list of excluded studies not provided;

- arbitrary threshold used to categorize the quality of included studies.

Maimaris, Paty, et al., 2013 [34]

 - Narrative review (systematic)

 - Individuals with hypertension

 - PLOS One

The influence of national or regional health systems on hypertension awareness, treatment, and control?

 - drug use: yes

 - healthcare use: no

 - health: yes

- a priori’ design: yes

- search comprehensive: yes

- grey literature: no

- year of last search: 2013

- # of studies included: total, 53; drugs/cost-sharing/ins, 35; Canada, 0

- duplicate study selection and data extraction: yes

Risk of bias for observational study designs were assessed using three domains: selection bias, information bias, and confounding. For RCTs, the Cochrane risk of bias tool was used;(5) Cochrane tool was not described or discussed. Only global ratings provided for both tools. Unclear how domains were operationalized and assessed.

- grey literature not searched;

- list of excluded studies not provided;

- preponderance of US studies limits the generalizability of the findings.

- formal quality assessment of included studies poorly described and discussed; only global ratings provided; unclear how any of the domains were operationalized and assessed.

Pimentel, Lapane, Briesacher, 2013 [35]

 - Narrative review (systematic)

 - US Medicare population (65+ years) in long-term care

 - Drugs Aging

The impact of US Medicare Part D on the long-term care context, specifically costs to long-term care residents, providers and payers; prescription drug coverage and utilization; and clinical and administrative outcomes.

 - drug use: yes

 - healthcare use: no

 - health: yes

- a priori’ design: no

- search comprehensive: yes

- grey literature: yes

- year of last search: 2013

- # of studies included: total, 19; drugs/cost-sharing/ins, 11; Canada, 0

- duplicate study selection and data extraction: no

Quality rating scale developed by Downs and Black (1998) to assess study quality.(6) Unclear which items were removed and which were kept. Some global ratings reported. Full assessment not provided. Unclear how any of the domains were operationalized and assessed.

- no ‘a priori’ design;

- list of excluded studies not provided;

- no duplicate study selection and data extraction;

- formal quality assessment of included studies poorly described and discussed; only global ratings provided; unclear how any of the domains were operationalized and assessed.

Sinnott, Buckley, et al., 2013 [36]

 - Meta-analysis (meta-analysis)

 - Publicly insured populations

 - PLOS One

The effect of copayments for prescriptions on adherence to prescription medicines in publicly insured populations.

 - drug use: yes

 - healthcare use: no

 - health: no

- a priori’ design: no

- search comprehensive: yes

- grey literature: yes

- year of last search: 2012

- # of studies included: total, 7; drugs/cost-sharing/ins, 7; Canada, 0

- duplicate study selection and data extraction: yes

Controlled before-and-after studies and interrupted time series designs were assessed using a modified version of the Cochrane EPOC criteria;(1) Cohort studies were assessed using the Effective Public Health Practice Project component rating scale.(4) Only global ratings provided. Unclear how domains were operationalized and assessed; 6/7 studies rated as weak, 1 study rated as weak/moderate.

- no ‘a priori’ design;

- quality assessment: only global ratings provided; unclear how any of the domains were operationalized and assessed;

- arbitrary threshold used to categorize the quality of included studies;

- small number of included studies limits the generalizability of the findings.

Kiil, Houlberg, 2014 [37]

 - Narrative review (systematic)

 - General population

 - European Journal of Health Economics

What is the extent to which copayment reduces individual demand for services on which it is imposed, has adverse health effects, and give rise to distributional consequences.

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: no

- grey literature: yes

- year of last search: 2011

- # of studies included: total, 47; drugs/cost-sharing/ins, 18; Canada, 9

- duplicate study selection and data extraction: unclear

None

- no ‘a priori’ design;

- list of excluded studies not provided;

- unclear duplicate study selection and data extraction;

- non-systematic search strategy;

- no formal quality assessment of included studies.

Mann, Barnieh, et al., 2014 [38]

 - Narrative review (systematic)

 - Individuals with cardiovascular-related chronic disease

 - PLOS One

The impact of drug insurance and varying levels of patient cost-sharing on medication adherence, clinical and economic outcomes in patients with cardiovascular-related chronic disease.

 - drug use: yes

 - healthcare use: no

 - health: yes

- a priori’ design: no

- search comprehensive: yes

- grey literature: no

- year of last search: 2013

- # of studies included: total, 11; drugs/cost-sharing/ins, 11; Canada, 3

- duplicate study selection and data extraction: yes

Cochrane risk of bias tool for RCT (5), and Cochrane EPOC (1) taxonomy for controlled before-after studies and interrupted time series designs were used to assess the quality of the studies. Seven components were rated as low, mid, high risk. Unclear how domains were operationalized and assessed.

- no ‘a priori’ design;

- list of excluded studies not provided;

- grey literature not searched;

- unclear how any of the quality criteria were operationalized and assessed;

- quality assessment not explicitly taken into account.

Kesselheim, Huybrechts et al., 2015 [39]

 - Narrative review (systematic)

 - General population

 - American Journal of Public Health

To determine how expansions or restrictions in prescription drug insurance have affected patients’ health outcomes or their use of health care services.

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: yes

- grey literature: no

- year of last search: 2014

- # of studies included: total, 23; drugs/cost-sharing/ins, 23; Canada, 0

- duplicate study selection and data extraction: yes

Used guidelines outlined in Cochrane Handbook for Systematic Reviews of Interventions (Cochrane risk of bias tool). Summary scores presented for each component and overall (low, unclear, high), 22/23 scored low/unclear. Unclear how domains were operationalized and assessed.

- no ‘a priori’ design;

- grey literature not searched;

- surprisingly low number of studies identified;

- list of excluded studies not provided;

- only summary scores presented; unclear what led to low quality scores;

- 22 of 23 included studied were conducted in the United States which limits the generalizability of the findings.

Luiza, Chavez et al., 2015 [40]

 - Narrative review (systematic)

 - General population

 - Report: The Cochrane Library

The effects of cap and co-payment policies on rational use of medicines, healthcare utilization, health outcomes and costs.

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: yes

- search comprehensive: yes

- grey literature: yes

- year of last search: 2013

- # of studies included: total, 32; drugs/cost-sharing/ins, 32; Canada, 9

- duplicate study selection and data extraction: yes

For RCTs, and Interrupted time series (ITS)/Repeated Measures (RM), assessed risk of bias using the Cochrane EPOC tool criteria that provides nine standard domains for RCTs and seven domains for ITS/RM. Summary ratings and justifications were provided for each study.(1) Confidence in overall estimates graded using GRADE.

- restrictive inclusion criteria limits the usefulness of the review.

Aziz, Hatah, et al., 2016 [41]

 - Narrative review (systematic)

 - General population

 - Patient Prefer Adherence

How do payment scheme affect patients’ adherence to medications.

 - drug use: yes

 - healthcare use: no

 - health: no

- a priori’ design: no

- search comprehensive: yes

- grey literature: no

- year of last search: 2015

- # of studies included: total, 21; drugs/cost-sharing/ins, 21; Canada, 2

- duplicate study selection and data extraction: no

27 items from Downs and Black’s checklist for measuring study quality was used.(6)

Only global ratings provided. Yes/No provided for all items, without any supporting information. Quality assessment not otherwise used.

- no ‘a priori’ design;

- no duplicate study selection;

- grey literature not searched;

- concerning low number of studies included;

- quality of the included studies not used in interpreting findings or formulating conclusions.

Banerjee, Khandelwal, et al., 2016 [42]

 - Narrative review (systematic)

 - Individuals with cardiovascular diseases

 - Open Heart

What are the barriers and facilitators to adherence to secondary cardiovascular disease prevention medications at health system level.

 - drug use: yes

 - healthcare use: no

 - health: no

- a priori’ design: yes

- search comprehensive: yes

- grey literature: yes

- year of last search: 2015

- # of studies included: total, 25; drugs/cost-sharing/ins, 4; Canada, 0

- duplicate study selection and data extraction: yes

Observational studies: three domains were assessed: selection bias, information bias and confounding. RCTs: the Cochrane risk of bias tool was used. Assessment not provided (only presence or absence of bias for each component provided along with generic total assessment [low, unclear, high]). 3/4 relevant studies assessed as low risk of bias but high risk of confounding and 1 study assessed as unclear risk of bias. Unclear how domains were operationalized and assessed.

- quality assessment: only summary scores presented; unclear what led to low quality scores;

- small number of relevant included studies limits the usefulness and generalizability of the findings.

- list of excluded studies was not provided.

Doshi, Li et al., 2016 [43]

 - Narrative review (systematic)

 - Individuals using specialty drugs

 - Am J Manag Care

What is the impact of cost-sharing on utilization of speciality drugs indicated for rheumatoid arthritis, multiple sclerosis, and cancer, and on use of nondrug medical services, health outcomes and spending?

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: unclear

- grey literature: no

- year of last search: 2014

- # of studies included: total, 19; drugs/cost-sharing/ins, 19; Canada, 0

- duplicate study selection and data extraction: yes

No formal quality assessment conducted. Limitations of included studies generally discussed.

- no ‘a priori’ design;

- non-systematic search strategy;

- grey literature not searched;

- list of excluded studies not provided;

- no formal quality assessment of included studies.

Powell, Saloner, Sabik, 2016 [44]

 - Narrative review (systematic)

 - Medicaid beneficiaries

 - Medical Care Research and Review

What are the effects of cost-sharing, focusing on low-income populations in US, on health care utilization, and spending

 - drug use: yes

 - healthcare use: yes

 - health: no

- a priori’ design: no

- search comprehensive: yes

- grey literature: yes

- year of last search: 2014

- # of studies included: total, unclear; drugs/cost-sharing/ins, unclear; Canada, 0

- duplicate study selection and data extraction: unclear

No formal quality assessment conducted. Natural experiments, studies with larger sample sizes (n > 1000), and studies that used representative samples from multiple states, considered of higher quality.

- no ‘a priori’ design;

- no/unclear duplicate study selection and data extraction;

- list of included and excluded studies not provided;

- results not clearly synthesized

- US focus limits the generalizability of the findings.

Gourzoulidis, Kourlaba, et al., 2017 [45]

 - Narrative review (systematic)

 - Individuals with heart failure or diabetes mellitus

 - Health Policy

To determine the association between copayment, medication adherence and outcomes in patients with heart failure and diabetes mellitus.

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: unclear

- grey literature: no

- year of last search: not reported

- # of studies included: total, 38; drugs/cost-sharing/ins, 11; Canada, 0

- duplicate study selection and data extraction: yes

A modified version of the EPHPP tool was used where the selection bias domain was replaced with allocation bias.(4) Only global ratings provided. Unclear how any of these criteria were operationalized and assessed.

- no ‘a priori’ design;

- no duplicate study selection and data extraction;

- grey literature not searched;

- list of excluded studies not provided;

- unclear how any of the quality criteria were operationalized and assessed;

- arbitrary threshold used to categorize the quality of included studies;

- limited generalizability of the findings (10/11 relevant studies used US data).

Park, Martin, 2017

 - Narrative review (systematic) [7]

 - Medicare Part D enrolees

 - Health Services Research

To update a a review on whether Medicare Part D changed drug utilization and out-of-pocket (OOP) costs overall and within subpopulations, and to identify evidence gaps.

 - drug use: yes

 - healthcare use: no

 - health: no

- a priori’ design: no

- search comprehensive: yes

- grey literature: yes

- year of last search: 2015

- # of studies included: total, 65; drugs/cost-sharing/ins, 62; Canada, 0

- duplicate study selection and data extraction: no

Yes; the authors used a predefined criteria and a fixed coding guide to systematically assess the risk of bias in each study. Rubric included bias common to observational studies: comparability between intervention and control groups, attrition, data collection and quality, measurement error, missing data, and reliability and validity of the outcome measures.

- no ‘a priori’ design;

- list of excluded studies not provided;

- no duplicate study selection and data extraction;

- US focus limits the generalizability of the findings.

Gupta, McColl et al., 2018 [46]

 - Narrative review (scoping)

 - General population (Canadians)

 - Patient Prefer Adherence

The extent, determinants, and consequences of cost-related nonadherence to prescription medications in Canada.

 - drug use: yes

 - healthcare use: yes

 - health: yes

- a priori’ design: no

- search comprehensive: yes

- grey literature: no

- year of last search: 2018

- # of studies included: total, 37; drugs/cost-sharing/ins, 20; Canada, 20

- duplicate study selection and data extraction: unclear

No formal quality assessment conducted. Limitations of included studies generally discussed.

- no ‘a priori’ design;

- no/unclear duplicate study selection and data extraction;

- grey literature not searched;

- list of excluded studies not provided;

- no formal quality assessment of included studies.

Ofori-Asenso, Jakhu et al., 2018 [47]

 - Meta-analysis (meta-analysis)

 - 65+ years statins users

 - Journals of Gerontology: Medical Sciences

What factors are associated with non-adherence and/or discontinuation of statins among older persons (65+)?

 - drug use: yes

 - healthcare use: no

 - health: no

- a priori’ design: yes

- search comprehensive: yes

- grey literature: yes

- year of last search: 2016

- # of studies included: total, 45; drugs/cost-sharing/ins, 6; Canada, 0

- duplicate study selection and data extraction: yes

Observational studies were assessed using a set of questions from the the National Institute of Health (NIH) Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies (unclear which questions were used).(7) RCTs were assessed using the Joanna Briggs Institute’s critical appraisal checklist for RCTs.(8) Tools not described or discussed. Only global ratings provided. Unclear how any of the domains were operationalized and assessed.

- grey literature not searched;

- list of excluded studies not provided;

- quality assessment: only global ratings provided; unclear how any of the domains were operationalized and assessed;

- arbitrary threshold used to categorize the quality of included studies.

Schneider, Gaedke et al., 2018 [48]

 - Meta-analysis (meta-analysis)

 - Individuals with chronic cardiovascular disease

 - Int J Clin Pract

Effect of characteristics of pharmacotherapy on non-adherence in chronic cardiovascular disease.

 - drug use: yes

 - healthcare use: no

 - health: no

- a priori’ design: no

- search comprehensive: yes

- grey literature: unclear

- year of last search: 2016

- # of studies included: total, 31; drugs/cost-sharing/ins, 17; Canada, 1

- duplicate study selection and data extraction: no

A reporting checklist, Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) was used.(9) Each of the 22 criteria received scores of 0 to 1, total scores then transformed into percentage. Full assessment not provided. Only total scores provided. Unclear how any of the criteria were operationalized and assessed.

- no ‘a priori’ design;

- list of excluded studies not provided;

- no duplicate study extraction;

- unclear if grey literature was searched;

- formal quality assessment of included studies conducted using a reporting checklist, poorly described and discussed; only global scored provided; unclear how any of the domains were operationalized and assessed.

Cheen, Tan et al., 2019 [49]

 - Meta-analysis (meta-analysis)

 - Individuals with any of 6 common chronic diseases

 - Int J Clin Pract

To assess prevalence of primary medication non-adherence in six common chronic diseases (asthma, chronic obstructive pulmonary disease, depression, diabetes mellitus, hyperlipidaemia, hypertension and osteoporosis), to identify and categorize factors associated with primary medication nonadherence; and to explore characteristics that contributed to heterogeneity between studies.

 - drug use: yes

 - healthcare use: no

 - health: no

- a priori’ design: no

- search comprehensive: yes

- grey literature: no

- year of last search: 2018

- # of studies included: total, 33; drugs/cost-sharing/ins, 8; Canada, 0

- duplicate study selection and data extraction: yes

The Cochrane risk of bias tool was used to assess clinical trials,(5) the Newcastle-Ottawa Scale for assessing cohort studies,(2) and the National Heart, Lung and Blood Institute Quality Assessment Tool for assessing cross-sectional studies (NIH).(7)

Unclear how domains were operationalized and assessed for Newcastle-Ottawa scale and the National Heart, Lung and Blood Institute Quality Assessment Tool.

- no ‘a priori’ design;

- grey literature not searched;

- list of excluded studies not provided;

- formal quality assessment of included studies poorly described and discussed;

- results may not be generalizable to other chronic conditions not included.

Kolasa, Kowalcyzk, 2019 [50]

 - Narrative review (systematic)

 - General population

 - Health Economics, Policy and Law

Association between prescription drug cost-sharing and health care consumption and health outcomes.

 - drug use: no

 - healthcare use: yes

 - health: yes

- a priori’ design: yes

- search comprehensive: yes

- grey literature: no

- year of last search: 2016

- # of studies included: total, 18; drugs/cost-sharing/ins, 18; Canada, 2

- duplicate study selection and data extraction: yes

Used a checklist, adapted from Gardner, Machin, Campbell (1986), for the assessment of the statistical content of medical studies. Tool not adequately described. Assessment not provided (only presence or absence of each component provided). Unclear how any of the domains were operationalized and assessed.(10)

- grey literature not searched;

- list of excluded studies not provided;

- arbitrary threshold used to categorize the quality of included studies;

- unclear how any of the quality criteria were operationalized and assessed;

- arbitrary threshold used to categorize the quality of included studies.

Mishuk, Fasina, Qian, 2019 [51]

 - Narrative review (systematic)

 - Individuals affected by US federal and state generic drug policies

 - Research in Social and Administrative Pharmacy

To evaluate the impact of US federal and state generic drug policies on drug use, spending, and patient outcomes.

 - drug use: yes

 - healthcare use: yes

 - health: no

- a priori’ design: no

- search comprehensive: yes

- grey literature: no

- year of last search: 2017

- # of studies included: total, 34; drugs/cost-sharing/ins, 24; Canada, 0

- duplicate study selection and data extraction: yes

The EPHPP tool was used to assess all included quantitative studies, however it is unclear if all eight domains were assessed or if a modified version of the tool was used. Only global ratings provided. Unclear how any of the domains were operationalized and assessed.(4)

- no ‘a priori’ design;

- grey literature not searched;

- list of excluded studies not provided;

- results not clearly synthesized;

- quality assessment: only global ratings provided; unclear how any of the domains were operationalized and assessed;

- US focus limits the generalizability of the findings.