Purpose
To summarize observational studies that focus on the use of glaucoma medications and to identify gaps in knowledge to guide future investigation.
Design
Literature study.
Methods
We searched the electronic databases MEDLINE, EMBASE, and PubMed for English language articles published through December 2009 using the search terms physician’s practice patterns , drug prescriptions , pharmaceutical services , medication adherence , ophthalmology , glaucoma , and ophthalmic solutions . We categorized studies by areas of focus and extracted and summarized key features: study population, data sources, and main findings.
Results
We identified 2224 articles by the search. Fifty-five described glaucoma medication use using large databases. Predominant areas of focus were: trends in prescription choices (n = 13); adherence, persistence, or both (n = 31); rational use of medications (n = 9); and policy-related issues (n = 2). Over the last decade, use of β-blockers and miotics has decreased substantially, whereas new agents, particularly prostaglandin analogs, have become more popular. Nonadherence was an issue in more than 25% of patients. A significant proportion of patients with comorbidities, contraindications, or both had received topical β-blockers.
Conclusions
To date, most studies have focused on adherence to glaucoma medications and changes in treatment choices. Major gaps in knowledge include prescribing patterns by prescriber specialty (ophthalmologists, primary care physicians, and optometrists), medication-related problems, and subsequent adverse health outcomes. Well-designed longitudinal observational studies addressing these gaps are warranted to improve patient safety.
Recently, nonphysicians are expanding their scope of practice into medicine, including ophthalmology. Optometrists in several countries, including the United Kingdom, Canada, and the United States, may prescribe some therapeutic agents. From January 2008, optometrists can prescribe a range of topical ophthalmic medications under Australia’s Pharmaceutical Benefits Scheme (PBS) after legislation passed by the Federal Parliament (National Health Amendment [Pharmaceutical Benefits] Act 2007 Commonwealth). These include lubricants and agents for treatment of allergy, infection, inflammation, and glaucoma. Extension of prescribing rights to optometrists under the PBS is likely to impact on the use of topical ophthalmic medications because the PBS supports universal access to prescription medicines subsidized by the commonwealth government. Eighty percent of prescription medicines used in the community are subsidized by the PBS.
Pharmacoepidemiology applies epidemiologic methods to assess use of pharmaceutical products and patient outcomes under nonexperimental situations. Pharmacoepidemiologic research uses various data sources, including patient questionnaires and medical and administrative records. Investigators frequently use large healthcare databases (e.g., pharmacy claims database), partly because recall and participation rates in studies using questionnaires can reduce the quality of information elicited. Data from general practitioners’ records and prescription claims databases often are representative and complete for very large patient populations and comprehensively cover different drug groups, patient populations, and geographical regions. Observational studies using large databases can describe patterns of medication use (e.g., selection of drug therapies, adherence) in clinical practice over time. These studies in large populations generally reflect real-world settings with respect to patient behavior and prescriber and other healthcare provider choices. The aims of the present study were: (1) to summarize the current state of population-based evidence on the use of glaucoma medications derived from observational studies analyzing electronic healthcare data, and (2) to identify gaps in knowledge to guide future investigation.
Methods
We assessed the literature for English language observational studies that focus on use of glaucoma medications. We searched MEDLINE (from 1950), EMBASE (from 1980), and PubMed (from 1949) through December 2009 using a combination of search terms describing drug use (e.g., drug utilization , physician’s practice patterns , drug prescriptions , pharmaceutical services , medication adherence ), and ophthalmology practice (e.g., ocular , ophthalmic , glaucoma ). The reference lists of studies identified in this way were screened to identify studies that our search strategy may have missed. We also searched by the names of first authors from the relevant papers.
A total of 2224 articles were identified by the search. Two authors (V.H.L. and C.Y.L.) examined these titles and abstracts for inclusion. We excluded nonoriginal articles and clinical studies investigating the efficacy, safety, or both, of a drug or drug group. We studied complete articles if they reported original research pertaining to glaucoma medicines, and only included articles if medication use was a main outcome measure. We excluded studies using questionnaires or medical chart review, focusing on population-based evidence as noted. We extracted and summarized key features (study population, data sources, outcome measures, main findings) of the selected studies.
Results
Fifty-five studies (56 publications between 1993 and 2009) used electronic healthcare data to assess use of glaucoma medications. We classified studies according to their predominant area of focus ( Table ), including: trends in prescription medication choices (n = 13); medication adherence, persistence, or both (n = 31); rational use of medications (n = 9); and policy-related issues (n = 2). Most studies were conducted in the United States (n = 31). More than half of the studies (n = 36) had large sample sizes as expected using computerized healthcare data. Studies of treatment trends analyzed aggregate data over time, whereas most other studies appropriately analyzed person-level data for compliance issues, medication-related problems, and impact of policy changes. Details on the 55 studies are provided as Supplemental Materials at AJO.com .
| Treatment Trends (n = 13) | Medication Adherence or Persistence (n = 31) | Rational Use of Medicines (n = 9) | Policy-Related Issues (n = 2) | |||
|---|---|---|---|---|---|---|
| Adherence (n = 9) | Persistence (n = 19) | Both (n = 3) | ||||
| Country | ||||||
| United States | 2 | 9 | 11 | 3 | 5 | 1 |
| United Kingdom | 1 | 7 | ||||
| Australia | 1 | 1 | 1 | |||
| Canada | 1 | |||||
| Israel | 1 | 2 | ||||
| Finland | 1 | |||||
| Others | 8 | |||||
| Sample size | ||||||
| <1000 subjects | 2 | 1 | 3 | 1 | ||
| 1000 to 5000 subjects | 1 | 2 | 9 | 2 | 1 | |
| 5000+ subjects | 1 | 5 | 7 | 7 | 1 | |
| Not clearly reported | 9 | 1 | 1 | 1 | ||
Treatment Trends
Thirteen studies (14 articles) included in this report examined trends in treatments for glaucoma or ocular hypertension in 11 countries. All studies had longer than 3 years of observation. Some studies (n = 5) examined trends specifically in patients with (or suspected of) glaucoma, providing more accurate results than those examining trends in total member populations ( Supplemental Table 1 ). Between 1994 and 2005, consistently across all countries, there was a dramatic shift from β-blockers and miotics to newer topical agents (α-agonists, carbonic anhydrase inhibitors, fixed combination products, prostaglandin analogs). Several studies also reported concurrent reductions in trabeculectomies and other glaucoma surgeries (from 17% in Canada to 67% in Scotland), particularly since availability of prostaglandin analogs.
Medication Adherence and Persistence
Thirty-one studies investigated patient adherence (n = 12) or persistence (n = 22) in glaucoma. Adherence is broadly defined as the extent to which patients take the medications as prescribed. Adherence outcome measures used by studies in this report included: duration without therapy over a certain period, interval between refills, and medication possession ratio (measuring the proportion of days on which the patient has the medication to use during a certain period). Persistence generally measures the time from drug initiation to discontinuation or change in treatment using survival analysis (Cox regression). Measures of persistence used by studies in this article included: time to discontinuation (discontinuation as defined by specified gap between prescription refills), time to change of index therapy (switching or adding another agent), time to initiating of surgery, and proportion of patients who persisted with initial medication. Supplemental Table 2 provides details of how adherence and persistence were measured in each study.
Rates of nonadherence ranged between 23% and 60% over 12 months. Rates of nonpersistence ranged between 30% and 95% at 1 year. Overall, studies reported higher adherence or persistence, or both, with prostaglandin analogs, particularly latanoprost, compared with drugs such as β-blockers and carbonic anhydrase inhibitors. However, approximately half of patients discontinuing prostaglandin therapy failed to restart any other topical therapy. Four studies that compared adherence between different prostaglandin analogs found better adherence with latanoprost, but with a higher average cost per patient per year. The addition of a second medication significantly increased refill intervals of the primary therapy. Poorer persistence and adherence was associated with depression, fewer visits to the ophthalmologist, multiple separate medication bottles, and younger age (40 to 49 years). Patients with ocular hypertension were less likely to adhere than patients with primary open-angle glaucoma, a more severe condition. Switching medications resulted from inadequate efficacy, adverse effects (e.g., hyperemia), or both. Glaucoma therapy with a carbonic anhydrase inhibitor containing fixed combination products seemed more persistent than those containing an α-agonist. Travoprost was more persistent than either dorzolamide plus timolol or latanoprost plus timolol combination products at a lower cost.
Rational Use of Medications
Nine studies ( Supplemental Table 3 ) investigated rational use of medicines, that is, “patients receive medications appropriate to their clinical needs, in doses that meet their own individual requirements, for an adequate period of time, and at the lowest cost to them and their community” ( http://www.who.int/medicines/ ). Five articles reported medication-related problems among patients treated with topical β-blockers, which may cause systemic adverse effects. Cosupply of topical and systemic β-blockers was common (20%). Approximately 8% to 25% of patients using topical β-blockers had concurrent prescriptions for medicines to treat obstructive pulmonary disease (OPD) or congestive heart failure. More than half (56%) of patients treated with β-blockers had some contraindication or precaution to the use of β-blockers (i.e., asthma or OPD, diabetes mellitus, cardiac arrhythmias, and heart failure), although this proportion was lower than that of patients taking newer glaucoma medications (brimonidine, dorzolamide, and latanoprost; 72%). Among patients with both glaucoma and OPD, use of topical β-blockers was high (62%), and most were noncardioselective β-blockers. Kaiserman and associates investigated the relationship of topical β-blockers with depression; they found no increased risk of depression associated with the use of topical β-blockers among an elderly managed care population.
Three studies assessed the rate of adding adjunctive agents to travoprost, bimatoprost, or latanoprost as primary therapies. Patients using travoprost or bimatoprost had significantly lower rates of adjunctive medication use compared with those starting on latanoprost monotherapy. Time until addition of an adjunctive agent was longest for travoprost monotherapy compared with bimatoprost and latanoprost.
Although it is usually mild and self-limiting, hyperemia is a common adverse event of prostaglandin analogs and an important factor associated with stopping or switching medications. Overall, costs nearly doubled in glaucoma patients who discontinued the initial therapy owing to hyperemia versus those who were hyperemia free, and latanoprost-treated patients had lower prescription and hyperemia-related follow-up costs than those initially treated with either bimatoprost or travoprost.
Policy-Related Issues
Two studies evaluated various policy-related issues pertaining to glaucoma medications ( Supplemental Table 4 ). Law and associates examined an automatic medication switch of all patients receiving latanoprost to bimatoprost, which was the preferred agent on the formulary of a nationwide health maintenance organization. Effective intraocular pressure control was maintained after the switch in the vast majority of patients; thus, a systematic pharmacy-level therapeutic substitution seemed feasible. Use of dorzolamide and latanoprost increased markedly after patient copayments for these new agents were reduced, with most new patients initiating within 2 months of copayment reduction.
Results
Fifty-five studies (56 publications between 1993 and 2009) used electronic healthcare data to assess use of glaucoma medications. We classified studies according to their predominant area of focus ( Table ), including: trends in prescription medication choices (n = 13); medication adherence, persistence, or both (n = 31); rational use of medications (n = 9); and policy-related issues (n = 2). Most studies were conducted in the United States (n = 31). More than half of the studies (n = 36) had large sample sizes as expected using computerized healthcare data. Studies of treatment trends analyzed aggregate data over time, whereas most other studies appropriately analyzed person-level data for compliance issues, medication-related problems, and impact of policy changes. Details on the 55 studies are provided as Supplemental Materials at AJO.com .
| Treatment Trends (n = 13) | Medication Adherence or Persistence (n = 31) | Rational Use of Medicines (n = 9) | Policy-Related Issues (n = 2) | |||
|---|---|---|---|---|---|---|
| Adherence (n = 9) | Persistence (n = 19) | Both (n = 3) | ||||
| Country | ||||||
| United States | 2 | 9 | 11 | 3 | 5 | 1 |
| United Kingdom | 1 | 7 | ||||
| Australia | 1 | 1 | 1 | |||
| Canada | 1 | |||||
| Israel | 1 | 2 | ||||
| Finland | 1 | |||||
| Others | 8 | |||||
| Sample size | ||||||
| <1000 subjects | 2 | 1 | 3 | 1 | ||
| 1000 to 5000 subjects | 1 | 2 | 9 | 2 | 1 | |
| 5000+ subjects | 1 | 5 | 7 | 7 | 1 | |
| Not clearly reported | 9 | 1 | 1 | 1 | ||
Treatment Trends
Thirteen studies (14 articles) included in this report examined trends in treatments for glaucoma or ocular hypertension in 11 countries. All studies had longer than 3 years of observation. Some studies (n = 5) examined trends specifically in patients with (or suspected of) glaucoma, providing more accurate results than those examining trends in total member populations ( Supplemental Table 1 ). Between 1994 and 2005, consistently across all countries, there was a dramatic shift from β-blockers and miotics to newer topical agents (α-agonists, carbonic anhydrase inhibitors, fixed combination products, prostaglandin analogs). Several studies also reported concurrent reductions in trabeculectomies and other glaucoma surgeries (from 17% in Canada to 67% in Scotland), particularly since availability of prostaglandin analogs.
Medication Adherence and Persistence
Thirty-one studies investigated patient adherence (n = 12) or persistence (n = 22) in glaucoma. Adherence is broadly defined as the extent to which patients take the medications as prescribed. Adherence outcome measures used by studies in this report included: duration without therapy over a certain period, interval between refills, and medication possession ratio (measuring the proportion of days on which the patient has the medication to use during a certain period). Persistence generally measures the time from drug initiation to discontinuation or change in treatment using survival analysis (Cox regression). Measures of persistence used by studies in this article included: time to discontinuation (discontinuation as defined by specified gap between prescription refills), time to change of index therapy (switching or adding another agent), time to initiating of surgery, and proportion of patients who persisted with initial medication. Supplemental Table 2 provides details of how adherence and persistence were measured in each study.
Rates of nonadherence ranged between 23% and 60% over 12 months. Rates of nonpersistence ranged between 30% and 95% at 1 year. Overall, studies reported higher adherence or persistence, or both, with prostaglandin analogs, particularly latanoprost, compared with drugs such as β-blockers and carbonic anhydrase inhibitors. However, approximately half of patients discontinuing prostaglandin therapy failed to restart any other topical therapy. Four studies that compared adherence between different prostaglandin analogs found better adherence with latanoprost, but with a higher average cost per patient per year. The addition of a second medication significantly increased refill intervals of the primary therapy. Poorer persistence and adherence was associated with depression, fewer visits to the ophthalmologist, multiple separate medication bottles, and younger age (40 to 49 years). Patients with ocular hypertension were less likely to adhere than patients with primary open-angle glaucoma, a more severe condition. Switching medications resulted from inadequate efficacy, adverse effects (e.g., hyperemia), or both. Glaucoma therapy with a carbonic anhydrase inhibitor containing fixed combination products seemed more persistent than those containing an α-agonist. Travoprost was more persistent than either dorzolamide plus timolol or latanoprost plus timolol combination products at a lower cost.
Rational Use of Medications
Nine studies ( Supplemental Table 3 ) investigated rational use of medicines, that is, “patients receive medications appropriate to their clinical needs, in doses that meet their own individual requirements, for an adequate period of time, and at the lowest cost to them and their community” ( http://www.who.int/medicines/ ). Five articles reported medication-related problems among patients treated with topical β-blockers, which may cause systemic adverse effects. Cosupply of topical and systemic β-blockers was common (20%). Approximately 8% to 25% of patients using topical β-blockers had concurrent prescriptions for medicines to treat obstructive pulmonary disease (OPD) or congestive heart failure. More than half (56%) of patients treated with β-blockers had some contraindication or precaution to the use of β-blockers (i.e., asthma or OPD, diabetes mellitus, cardiac arrhythmias, and heart failure), although this proportion was lower than that of patients taking newer glaucoma medications (brimonidine, dorzolamide, and latanoprost; 72%). Among patients with both glaucoma and OPD, use of topical β-blockers was high (62%), and most were noncardioselective β-blockers. Kaiserman and associates investigated the relationship of topical β-blockers with depression; they found no increased risk of depression associated with the use of topical β-blockers among an elderly managed care population.
Three studies assessed the rate of adding adjunctive agents to travoprost, bimatoprost, or latanoprost as primary therapies. Patients using travoprost or bimatoprost had significantly lower rates of adjunctive medication use compared with those starting on latanoprost monotherapy. Time until addition of an adjunctive agent was longest for travoprost monotherapy compared with bimatoprost and latanoprost.
Although it is usually mild and self-limiting, hyperemia is a common adverse event of prostaglandin analogs and an important factor associated with stopping or switching medications. Overall, costs nearly doubled in glaucoma patients who discontinued the initial therapy owing to hyperemia versus those who were hyperemia free, and latanoprost-treated patients had lower prescription and hyperemia-related follow-up costs than those initially treated with either bimatoprost or travoprost.
Policy-Related Issues
Two studies evaluated various policy-related issues pertaining to glaucoma medications ( Supplemental Table 4 ). Law and associates examined an automatic medication switch of all patients receiving latanoprost to bimatoprost, which was the preferred agent on the formulary of a nationwide health maintenance organization. Effective intraocular pressure control was maintained after the switch in the vast majority of patients; thus, a systematic pharmacy-level therapeutic substitution seemed feasible. Use of dorzolamide and latanoprost increased markedly after patient copayments for these new agents were reduced, with most new patients initiating within 2 months of copayment reduction.
Discussion
This study assessed the current state of population-based evidence on the use of glaucoma medications derived from observational studies analyzing computerized healthcare data. It revealed that the state of the science pertaining to rational use of glaucoma medications or the impact of policy-related changes on glaucoma medication use is not well developed.
Studies identified a swing toward new prescription medications, particularly fixed combination products and prostaglandin analogs, between 1994 and 2005. Reductions were noted in glaucoma surgery rates.
The major focus of the published literature was adherence to and persistence with glaucoma therapies using rates of prescription refills (56%; n = 31). Our finding of poor adherence or persistence with glaucoma medications based on studies using computerized healthcare data is consistent with previous research that reviewed database studies as well as chart reviews, qualitative interviews, questionnaires—which are subject to recall and sampling biases—and electronic monitoring. Overall, patients tend to report higher adherence rates than those derived from pharmacy claims analysis, suggesting an overestimate of their own behavior. However, analyses using prescription claims data have important limitations. For example, misidentifying newly treated patients, misclassifying addition versus switching of medications, and lack of clinical information such as intraocular pressure control.
Major barriers to medication adherence or persistence identified in the literature include situational and environmental factors (49%) and medication regimen (32%); these include side effects and difficulty with potentially complex regimen of topical or oral medications, or both. Poor adherence to therapy for any chronic condition is a major healthcare issue; it may require changes in medication regimen, it may necessitate surgical intervention, it may adversely affect health outcomes, and it may influence healthcare expenditure. The following may have a positive impact on adherence: simplifying eye drop regimens, reminder devices, education, and care planning. However, definitive conclusions about the effectiveness of the above interventions cannot be drawn because of inadequate methodologies and heterogeneity of study design.
Drug–disease or drug–drug interactions are important issues in the rational use of medicines. For patients with glaucoma or ocular hypertension, the use of topical β-blockers potentially may exacerbate conditions such as OPD and bradyarrhythmias. Several studies in this report found a concerning proportion of patients receiving topical β-blockers who had 1 or more contraindication(s) or received systemic β-blockers concurrently. Changes in pharmacologic management or close monitoring may be required in patients thus affected.
This study has several limitations. First, our search strategy may have missed articles despite efforts to identify all of the published studies pertaining to use of glaucoma medications. Second, our report did not summarize trends in surgery rates because only studies that measured medication use were included. Third, because we focused on observational studies using large databases, reasons for nonadherence were not collated comprehensively. Finally, we did not assess the methodologic quality of these studies (e.g., appropriateness of control for confounding). Clear reporting of studies enables critical appraisal; researchers are encouraged to consult the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement to ensure good reporting of observational studies.
This investigation identified several major gaps in knowledge. Our understanding of the use patterns of glaucoma medications in countries other than the United States is limited. The majority of studies (23/31) on glaucoma medication adherence and persistence were conducted in the United States, and most evaluated patients with third-party pharmacy benefits. Second, although interventions such as mandatory therapeutic substitution and patient copayment increases have been used widely by healthcare payers internationally to control medication use and costs, the extent of their use for ophthalmic medications and their impacts have not been studied systematically. Finally, optometrists have been authorized to prescribe topical ophthalmic treatments in several countries, but little is known about their prescribing patterns. Well-designed longitudinal observational studies are needed to examine prescribing patterns by prescriber specialty (ophthalmologists, primary care physicians, and optometrists), the prevalence and type of medication-related problems (including nonadherence), and subsequent patient outcomes.
In conclusion, use of newer glaucoma medications has increased substantially over the last decade across countries. Poor adherence to and persistence with ocular hypotensive therapy is a significant healthcare issue. Development and rigorous evaluation of interventions to improve medication adherence and persistence are essential. Future research should address the scientific gaps identified, particularly analysis of prescribing patterns and medication-related problems by prescriber specialty.
Dr Christine Lu is supported by an Australian National Health and Medical Research Council Public Health Training Fellowship (Grant no. 456438 ). Dr Goldberg is a member of the advisory boards for Alcon, Allergan, Merck, and Pfizer and has received research support from Alcon , Allergan , and Pfizer . The remaining authors indicate no financial conflict of interest. Involved in literature search (V.H.L., C.Y.L.); Data collection (V.H.L., C.Y.L.); Management (V.H.L., C.Y.L.), analysis (V.H.L., C.Y.L.), and interpretation (V.H.L., I.G., C.Y.L.) of data; Preparation of manuscript (V.H.L., I.G., C.Y.L.); Review of manuscript (V.H.L., I.G., C.Y.L.); and Approval of manuscript (V.H.L., I.G., C.Y.L.). This literature study did not require approval from a research ethics committee.
Appendix
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
