Diabetes mellitus is a chronic metabolic disorder posing a significant economic burden on the US healthcare system. According to the American Diabetes Association, an estimated 22.3 million people in the United States were diagnosed with diabetes in 2012, representing approximately 7% of the population.1 The prevalence of diabetes increased by 23% from 2007 to 2012 and is projected to increase to 1 in 3 adults by 2050.1 In 2012, the total estimated cost of diagnosed diabetes was $245 billion, including $176 billion in direct medical costs and $69 billion in lost productivity.1
The most common long-term complications of diabetes mellitus are retinopathy, with a potential loss of vision; nephropathy leading to renal failure; peripheral neuropathy, which is associated with the risk for foot ulcers and amputation; and cardiovascular-related morbidity and mortality.2 Patients with diabetes often progress toward numerous metabolic abnormalities, leading to a high risk for cardiovascular-related morbidity and mortality, with greater disease severity associated with higher risk.3
Diabetes is classified into type 1, type 2, and gestational disease. Type 1 diabetes accounts for approximately 5% to 10% of all cases of diabetes in the United States, whereas type 2 diabetes accounts for 90% to 95% of all cases.2
Glycemic control is crucial for preventing or minimizing the long-term complications associated with diabetes. To achieve and maintain optimal glycemic control, type 1 diabetes is generally managed through lifestyle changes. Similarly, type 2 diabetes management may also require lifestyle changes (including diet), but the disease may progress to require a combination of oral medications, noninsulin injectables, and/or insulin therapy in addition to lifestyle changes.4
The American College of Endocrinology (ACE) and American Association of Clinical Endocrinologists (AACE) suggest lifestyle management for all phases of type 2 diabetes, intensifying at higher hemoglobin (Hb) A1c levels. The ACE/AACE guidelines also recommend initiating oral antidiabetic medications when the HbA1c level is between 6% and 7%, and adding insulin therapy when the HbA1c level exceeds 8% among therapy-naïve patients, typically beginning with basal (ie, long-acting) insulin, and adding bolus (ie, short-acting) insulin if further intervention is needed.4
Because of the substantial human and economic burdens of type 2 diabetes, there is interest in understanding real-world patient adherence to, and persistence with, insulin therapy in this patient population; adherence measures the use of a medication as directed during treatment, and persistence measures treatment duration.5 Previous research has described poor adherence to oral medications and to insulin therapy.6 Similarly, insulin persistence is low, ranging from 26% to 52% in the year after the initiation of basal insulin, and even lower, at 19% to 42%, for bolus insulin.7
Recent research suggests that patients with type 2 diabetes who start therapy or are converted to insulin therapy with a pen demonstrate comparable or improved medication adherence versus patients who receive insulin with a vial or syringe.8-10 Health resource utilization, based on claims for hypoglycemic events, emergency department visits, physician visits, and annual medication costs, was found to be lower in patients using insulin pens.8-10 Compared with syringes, insulin pen devices have been shown to provide more reliable, accurate, and simplified dosing.11-13
Insulin delivery systems other than a vial or a syringe have the potential to improve factors such as patient treatment satisfaction, treatment adherence, and clinical outcomes.9 The use of these systems, such as prefilled insulin pens, in the United States has lagged behind other countries.9
The substantial and growing burden of type 2 diabetes and opportunities to curb its associated costs have been the focus of policymakers, payers, and nonprofit organizations. Strategies to improve medication adherence and its potential to lower healthcare resource utilization and costs for patients with type 2 diabetes are of interest to a wide variety of stakeholders.14,15 Consequently, there is significant interest in understanding the association between insulin adherence and healthcare costs for patients with type 2 diabetes who are insulin pen users.
The objectives of this study were to determine if higher insulin pen adherence among patients with type 2 diabetes who are insulin pen users was associated with lower healthcare costs, and to describe the overall healthcare costs of patients with type 2 diabetes. This study may provide insights to payers and providers to guide future analyses in identifying ways to improve diabetes care outcomes and to lower the associated healthcare expenditures.16
Methods
Study Design
In this retrospective claims database study we analyzed privately insured patients diagnosed with type 2 diabetes between January 2006 and September 2010. In this study, we evaluated the impact of adherence to insulin therapy on healthcare costs among patients with type 2 diabetes using insulin pens.
For this study, we used 2 MarketScan research databases from Truven Health—the Commercial Claims and Encounters (commercial) database and the Medicare Supplemental and Coordination of Benefits (Medicare supplemental) database. The commercial database contains the inpatient, outpatient, emergency department, and outpatient prescription drug experiences of several million individuals and their dependents in the United States. The overall database includes individuals from more than 100 self-insured employers and health plans. The Medicare supplemental database contains the healthcare data of individuals with Medicare supplemental insurance paid for by employers. The MarketScan research databases contain the healthcare data of privately insured individuals covered under a variety of fee-for-service, fully capitated, and partially capitated health plans. The health plans include preferred provider organizations, point of service plans, indemnity plans, and HMOs.
Inclusion Criteria
The Truven Health MarketScan Research Databases were used to identify adults (aged ≥18 years) with at least 1 insulin pen prescription claim (ie, the index event) between January 2006 and September 2010. They were also required to have continuous medical and pharmacy benefits for 12 months before and after the index event. Patients were required to have a diagnosis of type 2 diabetes mellitus (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] code 250.x0 or 250.x2) and the use of any oral antidiabetes agent in the 12 months before the index event. Patients were excluded if they had a diagnosis of type 1 diabetes mellitus (ICD-9-CM code 250.x1 or 250.x3) or gestational diabetes (ICD-9-CM code 648.8x), or if they used an insulin pump or oral or inhaled insulin during the study period.
The demographic characteristics were defined at the index event, including age, sex, geographic region, and insurance plan type, as shown in Table 1. The clinical characteristics were defined separately for the 12 months before and after the index event based on the presence of nondiagnostic or nonancillary claims for microvascular conditions (ie, diabetic retinopathy, macular edema, diabetic neuropathy, amputation, ulceration, renal disease), macrovascular conditions (ie, myocardial infarction, ischemic heart disease, congestive heart failure, peripheral vascular disease, cerebrovascular disease), and other general comorbid conditions (ie, anxiety, depression, dyslipidemia, hypertension).
The use of biguanides, sulfonylureas, meglitinides, thiazolidinediones, alpha-glucosidase inhibitors, fixed-dose therapies, antihyperlipidemics, antihypertensives, antidepressants, antiobesity medications, antiemetic/antinausea medications, exenatide, and liraglutide was also reported separately in the 12 months before and after the index event.
Measurements
The Charlson/Deyo Comorbidity Index, which measures the severity of comorbid conditions, was calculated and reported separately for the 12 months before and after the index insulin pen claim.17 Adherence to insulin was measured using the medication possession ratio (MPR), which is a standard measure for assessing treatment adherence as the extent to which a patient acts in accordance with the prescribed dosing interval and dose of a regimen. Persistence is defined as the duration of time from the initiation of a therapy to its discontinuation. In this analysis, adherence is used as an overarching term to describe adherence to therapy and persistence with therapy. MPR is a frequently used measure of adherence.18 Because of the data source, however, MPR is actually measuring refill adherence, because the data source does not contain information on whether the medication was actually used by the patient.
The MPR was calculated using the days’ supply filled of the insulin prescription claims, which was adjusted to account for variations in time between insulin refills.19 The MPR was reported as a continuous measure. The patients were also stratified by MPR quintile, ranging from least compliant (MPR, 0-0.20) to most compliant (MPR, 0.81-1.00). Insulin nonpersistence was previously defined as the presence of a 90-day gap in prescription claims for insulin.20,21
The annual direct medical costs were calculated by summing the patient and payer portions of all health insurance claims for the year before and the year after the index claim. The total costs were reported, as were the inpatient, outpatient, emergency department, and outpatient pharmacy costs. The all-cause and diabetes-related costs before and after the index claim were reported separately. The diabetes-related costs were defined as the paid amount (health plan and payer portions) on claims with a primary diagnosis of type 2 diabetes in any setting of care. All costs were adjusted to 2011 US dollars using the medical component of the Consumer Price Index.
This was a descriptive study and, as such, standard descriptive tests were used, where appropriate, to identify significant differences across the MPR categories; Fisher’s exact tests were used for comparing the means between the cohorts, and chi-square tests were used for tests of proportions. The cohort demographics, resource utilization before and after the index claim, and the expenditures were compared across the MPR categories.
Results
Patient Demographics
A total of 32,361 patients met the study criteria. The average patient age was 59.1 years (standard deviation [SD], 11.6) years, and 52.4% of the patients were male. Of the sample, 97.1% had a capitated, noncapitated, or fee-for-service health plan. As shown in Table 1, the MPR does not appear to differ by region or plan type; however, the MPR does appear to differ by age, with the highest MPR among patients aged 55 to 64 years.
Clinical Characteristics
Overall, the average sample MPR was 0.63 (SD, 0.29). Table 2 summarizes the baseline clinical characteristics of the study population, stratified by MPR. The baseline Charlson/Deyo comorbidity score was 2.0 (SD, 1.7).
The most common comorbidities included hypertension (40.9%), dyslipidemia (27.8%), ischemic heart disease (19.9%), diabetic neuropathy (14.7%), and renal disease (15.9%); 5.1% of patients had a claim with an ICD-9-CM code for obesity. The patients frequently continued to fill oral antidiabetes medication prescriptions after initiating an insulin pen, with the most common being biguanides (54.9%) and sulfonylureas (48.8%).
Healthcare Costs
The patients’ expenditures before and after the index claim were analyzed by MPR quintiles, wherein patients were stratified from least adherent (MPR, 0-0.20) to most adherent (MPR, 0.81-1.00). Table 3 summarizes the preindex annual healthcare expenditures of the study patients.
The mean preindex all-cause annual per-patient expenditures totalled $19,612 (SD, $40,571). The mean preindex diabetes-related annual per-patient expenditures totaled $2866 (SD, $5187). The preindex outpatient annual per-patient pharmacy costs were higher for the most adherent patients compared with the least adherent patients ($5683 vs $3852, respectively); the outpatient pharmacy costs also accounted for a larger proportion (31%) of the preindex total costs among the most adherent patients versus the least adherent patients (14%).
In the preindex period, the mean all-cause annual per-patient expenditures in the least adherent group were $27,707 (SD, $53,270), whereas the mean all-cause expenditures were $18,068 (SD, $38,504) in the most adherent group, or 1.53 times (P <.001) higher in the least adherent subgroup.
Table 4 summarizes the postindex annual per-patient healthcare expenditures. The mean all-cause annual per-patient expenditures for insulin pen users during the study period were $24,680 (SD, $44,005). The mean diabetes-related annual per-patient expenditures totaled $4952 (SD, $5209) and significantly increased after the index for all MPR groups, except the least adherent. For the least adherent group, we observed that the inpatient costs were 29% of the total all-cause expenditures versus 19% of the total all-cause expenditures for the most adherent group, which are likely driven by differences in the proportion of patients with an inpatient stay (37.3% vs 25.3%; P <.001).
The postindex outpatient annual per-patient pharmacy costs were higher for the most adherent patients than for the least adherent patients ($10,174 vs $5395, respectively; P <.001); the outpatient pharmacy costs also represented a larger proportion of the total postindex healthcare costs among the most adherent patients (43%) compared with the least adherent patients (21%). In the postindex period, the total all-cause annual per-patient expenditures were 9.4% (P = .007) lower for the most adherent group ($23,839; SD, $33,617) than for the least adherent group ($26,310; SD, $49,026).
Discussion
This study evaluated the relationship of insulin adherence to healthcare costs for a population of patients with type 2 diabetes using an insulin pen. Poor adherence is of particular concern in patients with type 2 diabetes, because previous studies have demonstrated that improved adherence to insulin therapy may substantially reduce the direct and indirect medical costs of type 2 diabetes for these patients.22,23 Poor adherence not only increases the risk for poor diabetes-related clinical outcomes, but it may also heighten the likelihood of concomitant renal and cardiovascular damage.24
In our analysis, patients with type 2 diabetes in the lowest quintile of adherence (by MPR) had total healthcare expenditures of >$26,000 annually compared with <$24,000 in the most adherent quintile, nearly a 10% difference that is highly statistically significant. These differences in total expenditures remain, despite the finding that outpatient pharmacy–related costs in the most adherent subgroup were much higher than for the least adherent quintile (>$10,000 vs ~$5400, respectively). These relationships strongly suggest benefits of improved adherence to prescribed medical therapies in the population with type 2 diabetes.
A systematic review of the literature on adherence and persistence of pharmacotherapeutic diabetes management found that patients frequently fail to comply with lifestyle management plans and treatment for diabetes, including insulin, noninsulin injectables, and oral hypoglycemic agents.6 The reasons for poor adherence are multifactorial, including communication between patient and provider, inadequate patient knowledge about antidiabetes medications, complex treatment regimens and their required follow-up, and insulin (and needle) resistance.25-28 In addition to being more user-friendly and convenient, insulin pens offer improved dose accuracy; superior portability; and easier, less painful injections than vial and syringe methods.29-31
The burden of poorly managed type 2 diabetes includes costs directly associated with type 2 diabetes as well as related conditions that are associated with poorly managed type 2 diabetes. As part of a large survey of US Medicare beneficiaries with self-reported diabetes, Stuart and colleagues found that greater medication adherence was significantly associated with lower medical costs and a reduced risk for hospitalization; each additional antidiabetes drug prescription was associated with a net $71 decrease in Medicare spending, which incorporates the cost of the prescription.32 These findings are substantively similar to our analysis, albeit in a more aged patient population (Medicare beneficiaries) and using older data (1997-2004).
According to a study conducted in 2012 by the American Diabetes Association examining the economic burden of diabetes care, the most influential factors in driving costs stem from the inpatient setting, from increased hospitalization rates and longer average lengths of stay.1 Together, these 2 factors accounted for more than 40% of the medical cost of diabetes.1 In our current analysis, the presence and length of inpatient stays were significant cost-drivers. The patients in the least adherent quintile had an average inpatient admission cost of $7543, which was $3058 greater than the cost in the most adherent quintile ($4485); the patients in the most expensive quintile (MPR, 21-40) had an inpatient cost ($8674) that was nearly twice that of the most adherent quintile.
According to research by the National Institute of Diabetes and Digestive and Kidney Diseases, each percentage point reduction in HbA1c reduced the risk for microvascular complications by 40%.33 In another study, 78% of diabetes-related inpatient costs, 47% of physician office visit costs, 82% of emergency department visit costs, and 52% of hospital outpatient costs were attributed to a combination of only a few diabetes-related medical and cardiovascular conditions.1 Similarly, in a 2011 review, Asche and colleagues reported that diabetic patients with an MPR of <80% have a 2.5 times greater risk for hospitalization related to diabetes than patients with an MPR of >80% (odds ratio, 2.53; 95% confidence interval, 1.38-4.64), a finding mirrored in this current analysis.34,35
Limitations
This analysis is subject to limitations inherent in the data source. First, it does not include socioeconomic status, anthropometric information, race or ethnicity, and mortality information. The data source also lacked data on clinical characteristics, such as HbA1c levels, body weight, and body mass index or obesity, all of which may impact diabetes management and healthcare costs. Likewise, hypoglycemia during a hospitalization may have contributed to the discontinuation of insulin and could not be accounted for as a result of ICD-9 coding limitations. Economic factors, such as barriers to obtaining pharmacy benefits, were also not accounted for.
The data for this study came mainly from large US employers and health plans, and therefore may not be generalizable to patients covered by other types of health plans, such as those with no insurance coverage or patients who are covered through the Veterans Affairs.
Finally, our observational study design does not allow any causal inferences to be made regarding our findings; this study is not designed to describe a direct causal relationship between insulin MPR and healthcare costs. It is instead intended to present descriptive analyses of insulin pen adherence and the potential association with healthcare costs. The presence of differences in preindex costs for patients with higher or lower insulin pen adherence underlies the complexity of the relationship between diabetes management and healthcare costs.
Conclusions
This study adds to a growing body of literature describing the burden of poor management of type 2 diabetes, including poor insulin adherence. Our study results are consistent with the published literature in describing the benefits of better insulin adherence in terms of healthcare costs. As expected, increased adherence was associated with increased outpatient pharmacy costs; previous research has demonstrated that the increased costs associated with increased adherence were more than offset by associated reductions in other medical and pharmacy costs.32
The average sample MPR for our study population was 0.63 (SD, 0.29), indicating that the average adherence to insulin was still relatively poor and continues to be a challenge for the successful management of diabetes. A plausible goal for MPR is at least 0.8, which was obtained by approximately 33% of patients in this analysis.
More research is needed to identify specific characteristics of better adherence as well as strategies or technologies that can lead to improved adherence among patients with type 2 diabetes. Further research should focus on the development of adherence measures that capture the effects of human and economic factors that can influence medication adherence. Furthermore, future analysis should focus on the interplay among insulin pen adherence and other antihyperglycemic therapies; this descriptive analysis focuses only on insulin pen adherence.
Funding Source
This study was funded by Becton Dickinson.
Author Disclosure Statement
Dr Bonafede is an employee of Truven Health Analytics, which was contracted by Becton Dickinson to conduct this study; Ms Chandran is, Ms Nigam was, and Ms Saltiel-Berzin is an employee of Becton Dickinson; Dr Hirsch is an employee of and holds stocks in Becton Dickinson, as well as in Merck; Ms Lahue is an employee of and holds stocks in Becton Dickinson.
Ms Chandran is Senior Director, Health Economics and Outcomes Research, Becton Dickinson; Dr Bonafede is Director of Outcomes Research, Truven Health Analytics, Cambridge, MA; Ms Nigam was Senior Analyst, Health Economics and Outcomes Research, Becton Dickinson, at the time of this study; Ms Saltiel-Berzin is World Clinical Education Specialist, Diabetes Care, Department of Medical Affairs, Becton Dickinson; Dr Hirsch is Worldwide Vice President, Diabetes Care, Department of Medical Affairs, Becton Dickinson; Ms Lahue is Worldwide Vice President, Health Economics and Outcomes Research, Becton Dickinson, Franklin Lakes, NJ.
References
1. American Diabetes Association. Economic costs of diabetes in the U.S. in 2012. Diabetes Care. 2013;36:1033-1046. Erratum in: Diabetes Care. 2013;36:1797.
2. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care. 2014;37(suppl 1):S81-S90. Erratum in: Diabetes Care. 2014;37:887.
3. Sarwar N, Gao P, Seshasai SRK, et al; for the Emerging Risk Factors Collaboration. Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease: a collaborative meta-analysis of 102 prospective studies. Lancet. 2010;375:2215-2222. Erratum in: Lancet. 2010;376:958.
4. Jellinger PS, Davidson JA, Blonde L, et al; for the ACE/AACE Diabetes Road Map Task Force. Road maps to achieve glycemic control in type 2 diabetes mellitus: ACE/AACE Diabetes Road Map Task Force. Endocr Pract. 2007;13:260-268.
5. Campbell RK. Recommendations for improving adherence to type 2 diabetes mellitus therapy—focus on optimizing insulin-based therapy. Am J Manag Care. 2012;18(3 suppl):S55-S61.
6. Cramer JA. A systematic review of adherence with medications for diabetes. Diabetes Care. 2004;27:1218-1224.
7. Asche CV, Shane-McWhorter L, Raparla S. Health economics and compliance of vials/syringes versus pen devices: a review of the evidence. Diabetes Technol Ther. 2010;12(suppl 1):S101-S108.
8. Davis SN, Wei W, Garg S. Clinical impact of initiating insulin glargine therapy with disposable pen versus vial in patients with type 2 diabetes mellitus in a managed care setting. Endocr Pract. 2011;17:845-852.
9. Rubin RR, Peyrot M. Factors affecting use of insulin pens by patients with type 2 diabetes. Diabetes Care. 2008;31:430-432.
10. Xie L, Zhou S, Wei W, et al. Does pen help? A real-world outcomes study of switching from vial to disposable pen among insulin glargine-treated patients with type 2 diabetes mellitus. Diabetes Technol Ther. 2013;15:230-236.
11. Asche CV, Luo W, Aagren M. Differences in rates of hypoglycemia and health care costs in patients treated with insulin aspart in pens versus vials. Curr Med Res Opin. 2013;29:1287-1296.
12. Goldstein HH. Pen devices to improve patient adherence with insulin therapy in type 2 diabetes. Postgrad Med. 2008;120:172-179.
13. Davis EM, Christensen CM, Nystrom KK, et al. Patient satisfaction and costs associated with insulin administered by pen device or syringe during hospitalization. Am J Health Syst Pharm. 2008;65:1347-1357.
14. Roebuck MC, Liberman JN, Gemmill-Toyama M, Brennan TA. Medication adherence leads to lower health care use and costs despite increased drug spending. Health Aff (Millwood). 2011;30:91-99.
15. Jha AK, Aubert RE, Yao J, et al. Greater adherence to diabetes drugs is linked to less hospital use and could save nearly $5 billion annually. Health Aff (Millwood). 2012;31:1836-1846.
16. Gibson TB, Song X, Alemayehu B, et al. Cost sharing, adherence, and health outcomes in patients with diabetes. Am J Manag Care. 2010;16:589-600.
17. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45:613-619.
18. Andrade SE, Kahler KH, Frech F, Chan KA. Methods for evaluation of medication adherence and persistence using automated databases. Pharmacoepidemiol Drug Saf. 2006;15:565-574; discussion 575-577.
19. Buysman E, Conner C, Aagren M, et al. Adherence and persistence to a regimen of basal insulin in a pre-filled pen compared to vial/syringe in insulin-naïve patients with type 2 diabetes. Curr Med Res Opin. 2011;27:1709-1717.
20. Bonafede MMK, Kalsekar A, Pawaskar M, et al. A retrospective database analysis of insulin use patterns in insulin-naïve patients with type 2 diabetes initiating basal insulin or mixtures. Patient Prefer Adherence. 2010;4:147-156.
21. Bonafede MM, Kalsekar A, Pawaskar M, et al. Insulin use and persistence in patients with type 2 diabetes adding mealtime insulin to a basal regimen: a retrospective database analysis. BMC Endocr Disord. 2011;11:3.
22. Wild H. The economic rationale for adherence in the treatment of type 2 diabetes mellitus. Am J Manag Care. 2012;18(3 suppl):S43-S48.
23. Balkrishnan R, Rajagopalan R, Camacho FT, et al. Predictors of medication adherence and associated health care costs in an older population with type 2 diabetes mellitus: a longitudinal cohort study. Clin Ther. 2003;25:2958-2971.
24. Akhrass F, Skinner N, Boswell K, Travis LB. Evolving trends in insulin delivery in pursuit of improvements in diabetes management. Am Health Drug Benefits. 2010;3:117-122.
25. Lin EHB, Ciechanowski P. Working with patients to enhance medication adherence. Clin Diabetes. 2008;26:17-19.
26. Osterberg L, Blaschke T. Adherence to medication. N Engl J Med. 2005;353:487-497.
27. Polonsky WH, Fisher L, Guzman S, et al. Psychological insulin resistance in patients with type 2 diabetes: the scope of the problem. Diabetes Care. 2005;28:2543-2545.
28. Tarn DM, Heritage J, Paterniti DA, et al. Physician communication when prescribing new medications. Arch Intern Med. 2006;166:1855-1862.
29. Cuddihy RM, Borgman SK. Considerations for diabetes: treatment with insulin pen devices. Am J Ther. 2013;20:694-702.
30. Asamoah E. Insulin pen—the “iPod” for insulin delivery (why pen wins over syringe). J Diabetes Sci Technol. 2008;2:292-296.
31. Pearson TL. Practical aspects of insulin pen devices. J Diabetes Sci Technol. 2010;4:522-531.
32. Stuart BC, Simoni-Wastila L, Zhao L, et al. Increased persistency in medication use by U.S. Medicare beneficiaries with diabetes is associated with lower hospitalization rates and cost savings. Diabetes Care. 2009;32:647-649.
33. Centers for Disease Control and Prevention. National diabetes fact sheet: national estimates and general information on diabetes and prediabetes in the United States, 2011. 2011. www.cdc.gov/diabetes/pubs/pdf/ndfs_2011.pdf. Accessed April 13, 2015.
34. Asche C, LaFleur J, Conner C. A review of diabetes treatment adherence and the association with clinical and economic outcomes. Clin Ther. 2011;33:74-109.
35. Lau DT, Nau DP. Oral antihyperglycemic medication nonadherence and subsequent hospitalization among individuals with type 2 diabetes. Diabetes Care. 2004;27:2149-2153.