Persistence and adherence with Latanoprost: A Danish register‐based cohort study in older patients with glaucoma

This study aims to assess the association between switching patterns and adherence/persistence in Danish patients over the age of 65, who started their first‐ever glaucoma treatment with latanoprost eye drops.

A total of 9.3 million prescriptions of latanoprost eye drops were dispensed in Denmark in 2021 (Schmidt et al., 2016).The number of redemptions on a prescription varies and depends on the prescriber.It is a common clinical practice in Denmark to prescribe one to three bottles at a time with redemptions sufficient for a year.This practice requires patients to visit their physician once a year to extend an existing prescription.
In March 2011, branded latanoprost, Xalatan®, (Pfizer Inc.) lost its patent protection (Kolko & Koch Jensen, 2017).The expiry of Xalatan®'s patent enabled the introduction of generic latanoprost eye drops in Denmark.This expansion of the pharmaceutical market has made it possible to switch between different latanoprost eye drops that differ in composition (e.g.inactive ingredients, chemical as well as physical properties such as pH, drop volume, buffer capacity, osmolality, surface tension, bottle colour and bottle design) (Chambers, 2012;Kolko & Koch Jensen, 2017).However, this does not apply to all latanoprost formulations.Monoprost ® (Laboratoires THEA, France) is a branded preservative-free latanoprost formulation in single-dose vials without a generic medicinal product in Denmark.The current clinical practice does not allow a switch from a preservative-free to a preservativecontaining medicine with the same active ingredient.
Multiple shifts can lead to patient confusion due to different bottles and packages (Kolko & Koch Jensen, 2017).Glaucoma patients are predominantly elderly and may rely on visual memory when handling their eye drops, frequent changes can put adherence at risk (De Natale & Gandolfi, 2013).Multiple shifts can lead to a number of problems, as varying physico-chemical properties may translate into differences in pharmacokinetics, pharmacodynamics, and tolerability (Cantor, 1997;Patel et al., 2013).These differences in turn could affect the efficacy and safety (Golan et al., 2015;Narayanaswamy et al., 2007).We hypothesized that experiencing multiple shifts may alter patients' adherence and persistence to pharmacological treatment due to increasing regimen complexity and side effects (Chang et al., 2013;Tsai et al., 2003).Adherence is the extent to which a patient acts in accordance with the prescribed regimen and persistence is the duration of time from initiation to discontinuation of treatment (Cramer et al., 2008).
The impact of multiple shifts on adherence and persistence has not yet been studied in Denmark.This is crucial considering that adherence and persistence to medication are important for successful glaucoma treatment in order to reduce the risk of visual impairment and loss of vision-related quality of life.We decided to evaluate switching patterns with the overall aim of assessing the impact of multiple shifts on adherence and persistence in patients who were naïve to latanoprost and over 65 years of age.

| Data sources
All Danish citizens are assigned a personal civil registration number at birth or immigration.The civil registration number is an identifier that can be used to merge Danish administrative registers and retrieve nationwide information regarding socio-demographic characteristics, claimed prescriptions, in-and out-patient hospital discharges, vital status and cause of death.In the present study, data were collected from the Danish National Prescription Registry, the Danish National Patient Register, The Danish Register of Causes of Death and the Danish Civil Registration System.In these registers, data were available for all patients.Information about patients' comorbidity was obtained by looking at their medical history in the Danish National Patient Register and medicine retrieved in community pharmacies.In Denmark, multiple prescriptions can be redeemed by the same patient and more than one medicinal product can be redeemed within the same medical prescription.

| Ethics
The study does not need ethical approval and patient consent because Danish register-based cohort studies are exempted.Patient records/information before the analysis was pseudonymized.The University of Copenhagen and Statistics Denmark (project number 707278) have appropriate data approval from the Regional Capital Area Data Protection Agency to facilitate the conduct of the present study.In particular, this study is part of a bigger project entitled 'Identification and risk minimization of potentially inappropriate prescriptions in patients aged 65 or older', for which Dr. Sessa has established a cohort of individuals aged 65 or older in Denmark.

| Study period and study population
The study period was from 1 January 2012 to 31 December 2018.The study population was composed of all Danish citizens aged ≥65 years who redeemed their first prescription ever of latanoprost during the study period.All individuals were using latanoprost for ocular hypertension.To be included in the study population, patients should not have been exposed to any other pharmacological treatment of glaucoma (Anatomical Therapeutic Chemical Classification, ATC-code: S01E) prior to latanoprost.This approach has been used to avoid mixing prevalent and incident new users which, in the epidemiological analysis are known to introduce biases (Lund et al., 2015).The study population was followed for 1 year from the date of their first redeemed prescription of latanoprost, and only patients still on latanoprost treatment at the end of the first year of follow-up were included in the final study population.At the end of the first year of the follow-up period, the patients included in the final study population were assigned to three different cohorts: (1) Switchers or rather patients exposed to two or more preserved eye drop products containing latanoprost as the active ingredient.(2) Non-switchers or rather patients who were only exposed to one brand of preserved eye drops containing latanoprost as the active ingredient (except Monoprost®).(3) Monoprost® or rather patients who redeemed Monoprost® as their first prescription of latanoprost.Being included in one cohort was mutually exclusive to being included in the other cohorts.

| Follow-up period
The date of inclusion in one of the three cohorts was defined as the index date.Patients were followed from the index date for 1 year until the end of the data coverage or censoring.Patients were censored at the development of the study outcomes, end of the follow-up period, permanent emigration or death.

| Outcomes
The primary outcomes were adherence and persistence to latanoprost treatment within 1 year from the index date.PDC was used as a measure of adherence (Forbes et al., 2018).To calculate PDC and assess persistence, we assigned a duration of supply to each redeemed prescription during the follow-up period.The duration of supply was calculated using the researcher-defined duration (Meaidi et al., 2021).By using this method, the expected end date of supply of each redeemed prescription was calculated as the calendar date when the patient runs out of posological units.This calculation has been performed under the assumption that the patient used the medicine according to the recommendations in the patient information leaflet.The end date of supply was used to assess how long each patient waited to redeem the next prescription.If this wait exceeded 90 days (i.e., gap) then the patient was considered as discontinuing the treatment.

| Covariates
Selected comedications/co-morbidities, sex, age, and equalized income were selected as covariates.Patients' equalized income was assessed the year before the index date.Operational definitions for study covariates are found in Table S1.

| Statistical analysis
The baseline characteristics of the three cohorts were compared using anova for continuous variables and Fisher's exact test for categorical variables.Study covariates described earlier were used to compute the propensity score for being switchers, non-switchers or Monoprost® users.The Cox regression model was used to compute the hazard ratio (HR) of latanoprost discontinuation for the three cohorts (the non-switchers cohort was chosen as the reference level).In the crude analysis, exposure and outcome were included as independent and dependent variables, respectively, in the statistical model.In the adjusted analysis, the propensity score was added as an independent continuous variable in the model.The propensity score was estimated using a multinomial logistic regression model (McCaffrey et al., 2013).The CMA6 constructor from the AdhereR package was used to compute latanoprost treatment persistence allowing the carryover effect within the observation window (Dima & Dediu, 2017).Smoothed cumulative incidence proportions curves for treatment discontinuation were drawn accounting for censoring.Grey's test was used to compare the cumulative incidence functions of the three cohorts.The pairwise Wilcoxon rank sum test was used to compare the PDC among cohorts with Holm adjustment for multiple comparisons.The null hypothesis was no difference in the PDC for the three cohorts.

| R E SU LT S
A total of 10 014 patients were included in the study population.Of these, 2321 patients were identified as nonswitchers, 6396 as switchers and 1297 as Monoprost ® users.The mean age of the study population was 75.4 years (standard deviation, SD 7.6 years) with a proportion of women of 55%.Highly prevalent comorbidities included uncomplicated hypertension (29.5%), cataract (25%), chronic lung disease (11%), atrial fibrillation (10.9%) and hypercholesterolemia (10.4%).Additional demographic and clinical characteristics of the 10 014 patients included in the study population are listed in Table 1.No differences in censoring during the follow-up period have been observed among groups (Figure 1).

| Persistence
In the unadjusted analysis, cumulative incidence curves were generated to compare cumulative incidence for study outcomes between the three cohorts.The cumulative incidence proportion curves diverge significantly after 240 days from the index date.Nonswitchers had a cumulative incidence of discontinuation of 19%, switchers had a cumulative incidence of 9% and Monoprost® users had a cumulative incidence of 4%.The one-year cumulative incidences show statistically significant differences in discontinuation among the three cohorts (p < 0.0001) with Monoprost® users being the cohort with the lowest risk of discontinuation.In the adjusted analysis, where possible competing risks were taken into account, the results were consequently consistent with the results of the cumulative incidences.had a 53% lower risk of treatment discontinuation compared to the reference group within 1 year following the first redemption of latanoprost in both unadjusted (HR 0.47; 95% CI: 0.41-0.53;p < 0.001) and adjusted (HR 0.47; 95% CI: 0.42-0.53;p < 0.001) analyses.Concerning the risk of treatment discontinuation, Monoprost® users compared with the reference group had a 78% lower risk of the above result in both unadjusted (HR 0.22; 95% CI: 0.17-0.28;p < 0.001) and adjusted (HR 0.22; 95% CI: 0.17-0.29;p < 0.001) analyses.

| DI SC US SION
This study provides real-world evidence on the impact of switching between different latanoprost eye drops on adherence and persistence in older glaucoma patients.The study found that patients identified as switchers showed greater medication adherence and persistence than those who had been identified as non-switchers.We expected a reduced adherence and persistence in switchers, due to a previously observed variation in tolerability among different latanoprost medications (Diagourtas et al., 2018;Narayanaswamy et al., 2007).Nevertheless, we did not find this result in a real-world setting.Due to the observational nature of our study, we can only assume that there are reasons for the increased adherence and persistence among switchers.One explanation is that with a chronic disease, such as glaucoma, the cost differences between medications become significant over time.Previous studies have identified medication costs as a barrier to adherence (Friedman et al., 2008;Stryker et al., 2010).Choosing the medication with the lowest cost when obtaining a prescription may be associated with improved adherence and persistence (Stein et al., 2015).Prices vary widely depending on the brand-name medication and its generic counterparts.At the time of evaluation, branded latanoprost, Xalatan® from Pfizer was approximately 2.5 times more expensive than its corresponding generic Latanoprost from Stada (Danish Pharmaceutical Information A/S, 2019b; Danish Pharmaceutical Information A/S, 2019a).Provided non-switchers were only exposed to Xalatan® eye drops the striking price difference is likely to be the  explanation behind increased treatment discontinuation in this particular cohort.It should be mentioned that public in Danish physicians is ranked among the highest in industrialized countries.Danes are confident in their physician's recommendations regarding treatment (Blendon et al., 2014).If physicians assure their patients that the generic latanoprost eye drops are the same, they may remain adherent despite experiencing bothersome side effects from multiple shifts.
The study showed that Monoprost® users had the greatest adherence and persistence among the three cohorts.Such a finding could support the assumption that preservatives pose a challenge when using anti-glaucomatous eye drops (Pisella et al., 2002).The side effects are manifested as ocular surface disease and are often accompanied by pain and discomfort.This is a significant source of poor adherence and may reduce patients' willingness to continue treatment (Thygesen, 2018).Preservative-free eye drops reduce ocular surface diseases and increase patient satisfaction with treatment, improving both quality of life and adherence to treatment (Economou et al., 2018;Thygesen, 2018).Preservative-free eye drops are frequently produced in single-dose vials posing a challenge for elderly patients with reduced manual dexterity.Our study finding indicates that this potential drawback does not present a noticeable risk in altering patients' adherence/persistence to Monoprost® (Jaenen et al., 2007).Although Monoprost® users had the greatest adherence and persistence, it should be noted that patients treated with single-dose vials will be reminded to redeem the next prescription in time as the number of single-dose vials decreases.This is not the case in the patients using multidose preserved latanoprost generics.
The adherence analysis found that patients aged 65 or older show great medical adherence, adherence estimates at 1 year ranged from 92% to 99%.A PDC of 92% compared to 99% corresponds to a patient missing 24 to 36 days of IOP-lowering eye drops per year compared to 3 to 4 days per year.Furthermore, the assessment of persistence reveals that some patients have periods of complete cessation of treatment for at least 90 days, but later restart the medication.These results are clinically relevant because glaucoma is a chronic condition and reductions in IOP delay progression (Bergeå et al., 1999).

| Strengths and limitations
Apart from strengths and limitations related to the observational nature of the study and its design (Thygesen & Ersbøll, 2014), the main limitation of our study is represented by the lack of information on the 'real use' of retrieved prescriptions.The Danish National Prescription Registry can only provide the extent to which medication prescriptions are redeemed.Researchers cannot obtain information on whether a patient is acting by the prescribed range and dose of a dosing regimen.Therefore, we had a strong assumption in our study, or rather that a redeemed prescription was used by the patient and that patients exhibiting higher PDCs were more likely to take the medication as prescribed.Another limitation of this study is that we did not account for the non-comparability of the number of drops that can be obtained from different types of bottles.Finally, as with all observational studies, we cannot rule out the presence of unmeasured confounders that may have introduced biases in our analysis.
Strengths include the use of the entire Danish population aged 65 or older which minimizes the risk of selection bias.In addition, other strengths include the use of reliable data sources such as Danish registers, the establishment of outcome with high internal/ external validity, and the ability to real-world evidence from routine clinical practice.
In conclusion, this study found increased adherence and persistence in latanoprost users among those who redeemed a preservative-free formulation (Monoprost ® ) or switched between different latanoprost formulations.

AC K NO W L E DGE M E N T S
We are thankful to Laboratoire THEA for being part of the scientific discussion Laboratoire THEA did not influence the results published or the decision to publish.
Baseline characteristics of the three cohorts.

F
I G U R E 1 Cumulative incidence curves.

F
U N DI NG I N FOR M AT ION Maurizio Sessa and Morten Andersen belong to the Pharmacovigilance Research Center, Department of Drug Design and Pharmacology (DDP), University of Copenhagen, supported by a grant from the Novo Nordisk Foundation (NNF15SA0018404).Prof. Morten Andersen has participated in research projects funded by AstraZeneca, H. Lundbeck & Novartis, Pfizer and Janssen, with grants paid to the institutions where he has been employed and has personally received fees from Medicademy, the Danish Pharmaceutical Industry Association, for leading and teaching pharmacoepidemiology courses.Miriam Kolko is a consultant for Abbvie and Thea Pharmaceuticals as well as a speaker for Abbvie, Thea Pharmaceuticals and Santen.Maurizio Sessa, Morten Andersen, and Miriam Kolko are all part of the Personalized Research Cluster at DDP.The remaining author has no conflicts of interest.