You are here

Using Information on Patient Adherence to Antipsychotic Medication to Understand Their Adherence to Other Medications

Jason Shafrin PhD
Alison R. Silverstein MPH
Joanna P. MacEwan PhD
Darius N. Lakdawalla PhD
Ainslie Hatch PhD
Felicia M. Forma BSc

INTRODUCTION

Mental illness affects, either directly or indirectly, almost half of the U.S. population during their lifetime,1 and approximately 4% of adults experience serious mental illness (SMI) each year.2 In 2014, the annual economic costs of bipolar disorder, major depressive disorder (MDD), and schizophrenia were $158 billion, $167 billion, and $217 billion, respectively, with more than half of all costs coming from indirect sources such as incarceration, homelessness, and comorbid conditions.3 Many patients with SMI also experience chronic physical comorbidities.47 Compared with the general population, the risk of cardiovascular disease is one-and-one-half times greater in people with depression and two to three times greater among people with schizophrenia or bipolar disorder.8 The prevalence of diabetes mellitus is also two to three times greater among people with schizoaffective or bipolar disorder.9 Patients with comorbid serious mental and physical illness often have to manage multiple medications for their mental health symptoms,9 and have the additional burden of managing medications for their physical comorbidities.1013

In addition to the high prevalence of comorbid conditions in patients with SMI, medication nonadherence frequently has a negative effect on patient-health outcomes and health-related quality of life1419 and exacerbates SMI symptoms. Further, among patients with type-2 diabetes (T2D) and hypertension, nonadherence is an independent risk factor for increased mortality19 and increased acute cardiovascular events.20 Patient nonadherence has also been associated with greater medical costs to both the individual and the health-care system overall.2125

To tackle the challenge of nonadherence, new digital technologies including electronic medication-adherence monitoring systems have been developed to more accurately track real-time adherence for oral2527 and injectable28 medications. In 2017, there were over 300,000 digital-health applications (“apps”); among those, 28% of disease-specific apps were developed to improve outcomes and adherence for patients with mental health and behavioral disorders.29 For patients with SMI who are prescribed medications to treat non-SMI comorbid conditions, however, existing technologies might be able to track adherence for only one of their prescriptions.27,30 Thus, despite the promise of digital technology, providers, payers, and caregivers may be skeptical of the benefits of apps aimed at measuring adherence to an SMI medication when many patients with SMI have multiple prescriptions for a variety of comorbid conditions.

To address this issue, we investigated the degree to which information on patients’ adherence to their antipsychotic medication is associated with their adherence to other medications. Our study provides preliminary evidence demonstrating that using effective digital technologies to measure adherence to atypical antipsychotics could improve adherence predictions for other oral medications for comorbid SMI, diabetes, and hypertension.

METHODS

We conducted a retrospective claims analysis using data from the Truven MarketScan® Commercial Claims and Encounters Database (Truven Health Analytics, Ann Arbor, MI), Medicare Supplemental Claims and Encounters Database, and Medicaid Multistate Medicaid Database, from January 1, 2013 through December 31, 2014. Patients aged 18 years and older were included in the sample if they had ≥ 1 prescription for a medication of interest between July 1 and December 31, 2013; an SMI diagnosis in the same period; and a diagnosis of diabetes or hypertension between January 1 and December 31, 2014. Patients had to be continuously enrolled in their insurance plan between January 1 and December 31, 2014. An SMI diagnosis was defined as having ≥ 1 inpatient claim or ≥ 2 outpatient claims with a diagnosis of schizophrenia (International Classification of Diseases, Ninth Revision [ICD-9] 295.x),28 bipolar disorder (ICD-9 296.0x-296.1x, 296.4x-296.8x),4,31 or MDD (ICD-9 296.2x, 296.3x, 311.x).32 Diabetes and hypertension diagnoses, respectively, were defined as having ≥ 1 inpatient claim or ≥2 outpatient claims with a diagnosis of diabetes (ICD-9 250.00, 250.01, 250.02, 250.03, 790.2x, 791.5, 791.6, V45.85, V53.91, V65.46)33 or hypertension (ICD-9 401.9, 401.0, 402.x, 403.x, 404.x, 405.01, 405.09, 405.11, 405.19, 405.91, 405.99, 437.2).33 Patients were excluded if their age was not reported in the data, if they had filled any antipsychotic medications by mail order, or if they had an index fill for clozapine. As consistent monitoring and physician supervision is required for clozapine use,34 it can affect prescription refill and adherence rates.35

Medications of interest were oral ones widely used for SMI, T2D, or hypertension. Atypical antipsychotics were identified using National Drug Codes and J codes, and the list of therapies compared was drawn from recommendations from the American Psychiatric Association Clinical practice guidelines.11,12,36 Five molecules from each of five SMI therapeutic classes were selected to represent the most commonly prescribed medications for treating schizophrenia, bipolar disorder, and MDD, based on the literature (including clinical-treatment guidelines)1113,37,38 and clinician feedback. With this approach, we examined patient adherence to mood stabilizers, selective serotonin reuptake inhibitors (SSRIs), serotonin-norepinephrine reuptake inhibitors (SNRIs), and dopamine norepinephrine reuptake inhibitors (bupropion only) for treating other SMIs. Diabetes and hypertension medications were selected according to clinical guidelines for the treatment of each disease,14,39,40 for which we included four to five of the most prescribed classes. The diabetes classes included biguanides, sulfonylureas, thiazolidinediones, and dipeptidyl-peptidase-4 inhibitors. Antihypertensive medications included thiazide diuretics, acetylcholine esterase inhibitors, angiotensin receptor blockers, calcium channel blockers, and beta blockers.

We calculated adherence based on the patient’s proportion of days covered (PDC), defined as the total day’s supply for prescriptions made over a 12-month period divided by 365. Following quality measures used within the Healthcare Effectiveness Data and Information Set (HEDIS), among other sources,41,42 we determined that patients were adherent to their antipsychotic if PDC was ≥ 80%.

The primary outcome of our study was the accuracy of patients’ adherence to their antipsychotic medication as an indicator of adherence to other SMI, anti-diabetes, or anti-hypertension medications. For instance, if a patient is adherent to both medications, or not adherent to both, we define atypical antipsychotic adherence as an accurate indicator for that patient. On the other hand, if a patient is adherent to one but not to the other medication, we define atypical antipsychotic adherence as an inaccurate indicator. More formally, accuracy is the sum of the true positives and true negatives divided by the entire sample. True positives represent patients who were adherent to both medications; true negatives represent patients who were not adherent to either medication. In this study, false positives are defined as patients who were adherent to their atypical antipsychotic but not to their other medication, and false negatives are defined as patients who were not adherent to their atypical antipsychotic but were adherent to their other medication.

Secondary outcomes were positive and negative predictive values, along with sensitivity and specificity. In our study, positive predictive value (PPV) describes patients who were adherent to both medications among those who were adherent to their atypical antipsychotic. Negative predictive value (NPV) describes patients who were not adherent to either medication among those who were not adherent to their atypical antipsychotic. Sensitivity is equal to the proportion of patients who were adherent to their atypical antipsychotic and their other medication among all those who were adherent to their other medication. Specificity is the proportion of patients not adherent to their atypical antipsychotic and their other medication among all those who were not adherent to their other medication.

We then conducted a t test to compare each calculated measure of accuracy, predictive value, and validity to 50%, to examine whether the measure was meaningfully different from a random chance of adherence to both medications.

RESULTS

The total sample included 129,614 patients, most of whom were covered by Medicaid and the lowest number of whom were covered by Medicare (commercially insured, n = 57,260; Medicare, n = 6,977; Medicaid, n = 65,377). The mean age of the full sample was 44.8 years, and 62.2% of participants were female. Medicare patients had the greatest number of comorbidities as measured by the Charlson Comorbidity Index (1.3 vs. 0.4 for commercially insured patients and 0.8 for Medicaid patients). Major depressive disorder was the most prevalent type of SMI across all payers (52.2% among commercially insured patients; 53.8% among Medicare patients; and 43.2% among Medicaid patients). Medicare patients had the greatest prevalence of diabetes (26.8% vs. 12.3% of commercially insured patients and 24.0% of Medicaid patients), and hypertension was most prevalent in Medicare patients with SMI (55.6% vs. 22.2% of commercially insured patients and 39.2% of Medicaid patients). The average PDC was 0.725 (standard deviation [SD] = 0.307), with 53.7% considered adherent (i.e., PDC ≥ 80%). The proportion of patients adherent to their atypical antipsychotic was lowest among those who were commercially insured (51.4%) and highest among those who were covered by Medicaid (58.0%). Patient characteristics are detailed in Table 1.

Overall, patients’ adherence to their antipsychotic was a good indicator of adherence to their other medication in approximately two of every three patients. Specifically, the median atypical adherence accuracy across all 23 medications considered was 67.4% (all, P < 0.001), with accuracy ranging from 55.5% for valproate (P < 0.001) to 71.0% for sertraline (P < 0.001; Figure 1). The accuracy of atypical antipsychotic adherence was slightly higher for SMI medications (median, 68.6%) compared to the accuracy indicating adherence to diabetes or hypertension medication (median, 64.7% and 66.5%, respectively). For example, if patients are taking an atypical antipsychotic for schizophrenia and metformin for T2D, the accuracy calculation tells us that approximately 67/100 patients will have the same adherence patterns to both medications: they will adhere to both or adhere to neither.

Patients with poor adherence to antipsychotic agents were especially likely to be nonadherent to their other medications; however, patients who did adhere to antipsychotics were less likely to be indicative of adherence to other medications. Specifically, median PPV for atypical antipsychotics was 61.7%, and ranged from 42.6% to 67.0% (all, P < 0.001; Figure 3). In other words, 42.6% to 67.0% of patients who were adherent to their antipsychotic agent were also adherent to their other medication. Median PPV was 61.6% within other SMI medications, 57.6% within diabetes medications, and 64.2% within hypertension medications. NPV was higher than PPV across almost all medication pairs, ranging from 61.5% to 88.1% with a median value of 74.6%, (all, P < 0.001; Figure 4). In other words, 61.5% to 88.1% of patients who were not adherent to their antipsychotics were also not adherent to the other medication class. Median NPV was higher among SMI medications (77.7%) than among diabetes medications (75.1%) or hypertension medications (69.7%). Median sensitivity across all pairs was 73.7%, ranging from 67.2% to 90.1% (all, P < 0.001; Figure 5), and median specificity was 60.0%, ranging from 37.7% to 68.5% (all, P < 0.001, except thiazolidinedione [P = 0.065]; Figure 6). Adherence accuracy, predictive value (PPV/NPV), and validity (sensitivity/specificity) measures are summarized in Figure 2.

To better understand the predictive value and validity metrics, we refer to our example in which adherence to atypical antipsychotic agents is used to predict adherence to metformin: metformin PPV was 61.7%, NPV was 74.6%, sensitivity was 78.5%, and specificity was 56.5%. These figures imply that approximately 62/100 patients who were adherent to their atypical antipsychotic drug were also adherent to metformin, and approximately 75/100 patients who were not adherent to their atypical antipsychotic were also not adherent to metformin. Regarding the validity metrics, approximately 79/100 patients who were adherent to metformin were also adherent to their atypical antipsychotic agent, and approximately 56/100 patients who were not adherent to metformin were not adherent to their atypical antipsychotic.

Although the 80% threshold can appear somewhat arbitrary, our findings were not sensitive to this threshold. We conducted a sensitivity analysis to test other PDC thresholds (i.e., 60%, 70%, and 90%), and the results were similar (Table 2). Indicator accuracy ranged from 67.4% to 68.5% across these different thresholds. As expected, PPV decreased when the PDC threshold increased (from 69.6% at a 60% adherence threshold to 56.0% at a 90% adherence threshold), whereas the reverse was true for NPV (increasing from 65.0% at a 60% adherence threshold to 80.0% at a 90% adherence threshold).

DISCUSSION

Adherence to atypical antipsychotic medications provided useful information on concurrent adherence to other medications. For 67.4% of patients, adherence to their atypical antipsychotic medication and to their other SMI, diabetes, or hypertension medication was identical during the same time period. This accuracy level represents a substantial improvement when compared with physicians’ estimates of medication adherence. One study found that physician estimates of patient adherence to antipsychotics matched the adherence measured in claims data 53% of the time.43 These results may be driven by the fact that many physicians have limited formal training in detecting medication nonadherence.44

Among patients with both physical and mental illnesses, nonadherence to SMI medications contributes to nonadherence to medications for physical ailments.18,19,44,45 One study of elderly patients in Italy found that those who were adherent to SSRIs or SNRIs were more likely to be adherent to other prescribed medications, particularly antipsychotics.19 Our results echo those findings, and further contribute to the literature as our patient population is from a different geographic region (U.S.-based) and includes both elderly and non-elderly individuals.

Our study suggests that disease-specific digital-monitoring technologies that provide more accurate information on patient adherence to antipsychotic medications can be useful at predicting their adherence to other medications. As NPV was higher than PPV across most treatments, atypical antipsychotic adherence can be considered a necessary condition for adherence to other medications. This finding has been replicated in other studies.24,46 Because SMI-medication adherence is important for patients’ optimal symptom stability and day-to-day functioning, it may be a precondition for adherence to their other prescription medications.47

A better understanding of the adherence patterns in patients who have both serious mental and physical illnesses is important not only for improving the health of these patients, but for ensuring that future medical complications resulting from nonadherence are avoided. Treatment for individuals living with multiple chronic conditions accounts for 66% of U.S. health care costs.47 A number of initiatives have been developed to improve the quality and efficiency of the care received by such patients, including the Multiple Chronic Conditions Research Network, which is conducting research to improve healthcare quality and delivery for these patients.48 In addition, the Patient-Centered Medical Home (PCMH) program promotes comprehensive and coordinated care.48 If these and other initiatives are able to collect accurate information on patient adherence to antipsychotics via new technologies, they may be able to better predict adherence to other medications. Better coordination of care and proactive outreach to assess medication adherence for physical illness may also improve quality of care for these often difficult-to-treat patients and potentially reduce costs. Previous research has shown that for patients with SMIs, having access to adherence information can lead to improved treatment choices49 as well as cost savings.50

Limitations

There were a number of limitations in our study. First, health-insurance claims data measure adherence imperfectly: they reflect the filling of a prescription covered by insurance but do not record medication ingestion, and can thus underestimate adherence. On the other hand, as medications purchased outside the insurance system are not included in the data, claims-based adherence measures can also overestimate medication adherence. Second, adherence was measured over a one-year period, which might not account for changing adherence patterns within the year or over a patient’s lifetime. Third, there was no control for the number of medications prescribed for each patient, i.e., their pill burden, which can substantially affect patients’ ability to adhere to a medication regimen across multiple diseases.51 Fourth, these results might not be generalizable to cohorts of patients with other diseases, patients prescribed medications or combinations of medications that were not considered in this study, or patients outside the U.S. However, the accuracy calculations were fairly stable across the medications and diseases included in our analysis. Fifth, this study highlights the associations but cannot prove a causal relationship between adherence conditions in the medications of interest. Finally, our case study relies on claims data to infer the utility of more accurate digital-based adherence in understanding patient adherence to other medications. Using disease-specific digital technologies could be more or less accurate for measuring adherence to a patient’s other comorbid conditions than is the case in our claims-based analysis. As more of these technologies are being used in the real world, further research is needed.

CONCLUSION

Information on patient adherence to atypical antipsychotic medications can help predict patient adherence to other SMI, antidiabetes, and antihypertension medications. As the NPV is higher than the PPV, adherence to an antipsychotic could be a necessary but not sufficient condition for adherence to other medications. This case study suggests that using digital technology to measure adherence more accurately could supply additional information to providers and payers to help predict adherence to other medications.

Figures and Tables

Atypical antipsychotic adherence-prediction accuracy by medication. Adherence is defined as proportion of days covered (PDC) ≥ 0.80. A 50% accuracy rate implies that the accuracy metric is no better than random chance. All accuracy values were statistically different from 50%, with P < 0.001 in all cases.

ACE = angiotensin-converting enzyme; ARB = angiotensin receptor blocker; DPP4 =dipeptidyl-peptidase-4 inhibitor; SMI = serious mental illness; SNRI = serotonin-norepinephrine reuptake inhibitor; SSRI = selective serotonin reuptake inhibitor

Ability of atypical antipsychotic adherence to predict adherence to other medications; accuracy, validity, and predictive metrics. Box plots represent the ability of atypical antipsychotic adherence metrics (PDC ≥ 0.80) to match adherence across the other drug classes considered. The box-plot whiskers extend to the minimum and maximum ability to match adherence across the 23 other drug classes considered in this study. A 50% accuracy implies that the metric is no better than random chance. Median accuracy, 67.4%; median PPV, 61.7%; median NPV, 74.6%; median sensitivity, 73.7%; median specificity, 60.0%.

NPV = negative predictive value; PDC = proportion of days covered; PPV = positive predictive value

Atypical antipsychotic adherence positive predictive value (PPV) by medication. Adherence is defined as proportion of days covered (PDC) ≥ 0.80. A 50% PPV implies that the predictive metric is no better than random chance. All values were statistically different from 50%, with P < 0.001 in all cases.

ACE = angiotensin-converting enzyme; ARB = angiotensin receptor blocker; DPP4 = dipeptidyl-peptidase-4 inhibitor; SMI = serious mental illness; SNRI = serotonin-norepinephrine reuptake inhibitor; SSRI = selective serotonin reuptake inhibitor

Atypical antipsychotic adherence negative predictive value (NPV) by medication. Adherence is defined as proportion of days covered (PDC) ≥ 0.80. A 50% NPV implies that the predictive metric is no better than random chance. All values were statistically different from 50%, with P < 0.001 in all cases.

ACE = angiotensin-converting enzyme; ARB = angiotensin receptor blocker; DPP4 = dipeptidyl-peptidase-4 inhibitor; SMI = serious mental illness; SNRI = serotonin-norepinephrine reuptake inhibitor; SSRI = selective serotonin reuptake inhibitor

Atypical antipsychotic adherence sensitivity by medication. Adherence is defined as proportion of days covered (PDC) ≥ 0.80. A 50% sensitivity implies that the validity metric is no better than random chance. All values were statistically different from 50%, with P < 0.001 in all cases.

ACE = angiotensin-converting enzyme; ARB = angiotensin receptor blocker; DPP4 = dipeptidyl-peptidase-4 inhibitor; SMI = serious mental illness; SNRI = serotonin-norepinephrine reuptake inhibitor; SSRI = selective serotonin reuptake inhibitor

Atypical antipsychotic adherence specificity by medication. Adherence is defined as proportion of days covered (PDC) ≥ 0.80. A 50% specificity implies that the validity metric is no better than random chance. All values were statistically different from 50%, with P < 0.001 in all cases, except thiazolidinediones, P = 0.065.

ACE = angiotensin-converting enzyme; ARB = angiotensin receptor blocker; DPP4 = dipeptidyl-peptidase-4 inhibitor; SMI = serious mental illness; SNRI = serotonin-norepinephrine reuptake inhibitor; SSRI = selective serotonin reuptake inhibitor

Patient Characteristics, By Payer Type

Commercial (n = 57,260) Medicare (n = 6,977) Medicaid (n = 65,337)
Age, mean (SD) 43.0 (13.4) 73.9 (9.9) 43.4 (13.2)
Female, n (%) 36,141 (63.1%) 4,670 (66.9%) 39,825 (61.0%)
Schizophrenia diagnosis, n (%) 4,857 (8.5%) 585 (8.4%) 23,078 (35.3%)
Bipolar disorder diagnosis, n (%) 22,803 (39.8%) 1,550 (22.2%) 25,118 (38.4%)
MDD diagnosis, n (%) 29,907 (52.2%) 3,753 (53.8%) 28,256 (43.2%)
Diabetes diagnosis, n (%) 7,593 (13.3%) 1,871 (26.8%) 15,693 (24.0%)
Hypertension diagnosis, n (%) 12,697 (22.2%) 3,880 (55.6%) 25,652 (39.2%)
Charlson comorbidities, mean (SD) 0.4 (0.8) 1.3 (1.4) 0.8 (1.0)
Drug abuse, n (%) 4,769 (8.3%) 179 (2.6%) 9,167 (14.0%)
Alcoholism, n (%) 3,132 (5.5%) 183 (2.6%) 4,768 (7.3%)
Atypical antipsychotic PDC (SD) 0.700 (0.315) 0.746 (0.301) 0.727 (0.308)
Patients with atypical antipsychotic PDC ≥ 80%, n (%) 29,435 (51.4%) 4,050 (58.0%) 36,131 (55.3%)

n = number; SD = standard deviation; MDD = major depressive disorder; PDC = proportion of days covered

Median Accuracy, Predictive Value, and Validity Measurements Across All Atypical-Drug Pairs Using Alternative Definitions of Adherence

Accuracy PPV NPV Sensitivity Specificity
PDC 60% 68.5% 69.6% 65.0% 80.0% 50.1%
PDC 70% 68.2% 66.8% 68.8% 77.3% 55.1%
PDC 80% 67.4% 61.7% 74.6% 73.7% 60.0%
PDC 90% 67.8% 56.0% 80.0% 70.2% 64.6%

PDC 80% is the baseline measure.

PPV = positive predictive value; NPV = negative predictive value; PDC = proportion of days covered

Author bio: 
Dr. Shafrin, Ms. Silverstein, Dr. MacEwan, and Dr. Lakdawalla are based at Precision Health Economics in Los Angeles, California; Dr. Hatch is in the Clinical Sciences Department at Otsuka America Pharmaceutical, Inc. in Princeton, New Jersey; and Ms. Forma is in the Health Outcomes Department at Otsuka Pharmaceutical Development & Commercialization, Inc. in Princeton.

Financial support for this study was provided by Otsuka Pharmaceutical Development & Commercialization, Inc. under contract to Precision Health Economics (PHE). Dr. Shafrin, Dr. Silverstein, and Dr. MacEwan are employees of PHE, and Dr. Lakdawalla is a consultant to PHE. Dr. Hatch is an employee of Otsuka America Pharmaceutical, Inc. Ms. Forma is an employee of Otsuka Pharmaceutical Development & Commercialization, Inc.

References

  1. Kessler RC, Berglund P, Demler O, et al. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry 2005;62;(6):593–602.10.1001/archpsyc.62.6.593
  2. National Institute of Mental Health. Mental illness February 2019; https://www.nimh.nih.gov/health/statistics/mental-illness.shtml#part_154788. Accessed April 24, 2019
  3. MacEwan JP, Seabury S, Aigbogun MS, et al. Pharmaceutical innovation in the treatment of schizophrenia and mental disorders compared with other diseases. Innov Clin Neurosci 2016;13;(7–8):17–25.
  4. Carney CP, Jones LE. Medical comorbidity in women and men with bipolar disorders: a population-based controlled study. Psychosom Med 2006;68;(5):684–691.10.1097/01.psy.0000237316.09601.88
  5. Dixon L, Postrado L, Delahanty J, et al. The association of medical comorbidity in schizophrenia with poor physical and mental health. J Nerv Ment Dis 1999;187;(8):496–502.
  6. Henderson DC. Schizophrenia and comorbid metabolic disorders. J Clin Psychiatry 2005;66;(suppl 6):S11–S20.
  7. Kilbourne AM, Brar JS, Drayer RA, et al. Cardiovascular disease and metabolic risk factors in male patients with schizophrenia, schizoaffective disorder, and bipolar disorder. Psychosomatics 2007;48;(5):412–417.10.1176/appi.psy.48.5.412
  8. Kilbourne AM, Cornelius JR, Han X, et al. Burden of general medical conditions among individuals with bipolar disorder. Bipolar Disord 2004;6;(5):368–373.10.1111/j.1399-5618.2004.00138.x
  9. De Hert M, Correll CU, Bobes J, et al. Physical illness in patients with severe mental disorders. I. Prevalence, impact of medications and disparities in health care. World Psychiatry 2011;10;(1):52–77.
  10. Sachs GS, Thase ME, Otto MW, et al. Rationale, design, and methods of the systematic treatment enhancement program for bipolar disorder (STEP-BD). Biol Psychiatry 2003;53;(11):1028–1042.
  11. American Psychiatric Association. Practice guideline for the treatment of patients with bipolar disorder (revision). Am J Psychiatry 2002;159;(suppl 4):S1–S50.
  12. Lehman AF, Lieberman JA, Dixon LB, et al. Practice guideline for the treatment of patients with schizophrenia, second edition. Am J Psychiatry 2004;161;(suppl 2):S1–S56.
  13. In: Gelenberg AJ, Freeman MP, Markowitz JC, et al. Practice Guidelines for the Treatment of Patients with Major Depressive Disorder 3rd edWashington, D.C: American Psychiatric Association. 2010;
  14. Kukreja S, Kalra G, Shah N, Shrivastava A. Polypharmacy in psychiatry: a review. Mens Sana Monogr 2013;11;(1):82–99.10.4103/0973-1229.104497
  15. Aikens JE, Piette JD. Longitudinal association between medication adherence and glycaemic control in Type 2 diabetes. Diabet Med 2013;30;(3):338–344.10.1111/dme.12046
  16. Dempe C, Jünger J, Hoppe S, et al. Association of anxious and depressive symptoms with medication nonadherence in patients with stable coronary artery disease [published online January 3, 2013]. J Psychosom Res 2013;74;(2):122–127.10.1016/j.jpsychores.2012.12.003.
  17. Fenn HH, Bauer MS, Altshuler L, et al. Medical comorbidity and health-related quality of life in bipolar disorder across the adult age span. J Affect Disord 2005;86;(1):47–60.10.1016/j.jad.2004.12.006
  18. Gentil L, Vasiliadis HM, Préville M, et al. Association between depressive and anxiety disorders and adherence to antihypertensive medication in community-living elderly adults [published online October 30, 2012]. J Am Geriatr Soc 2012;60;(12):2297–2301.10.1111/j.1532-5415.2012.04239.x
  19. Weidenbacher HJ, Beadles CA, Maciejewski ML, et al. Extent and reasons for nonadherence to antihypertensive, cholesterol, and diabetes medications: the association with depressive symptom burden in a sample of American veterans. Patient Prefer Adherence 2015;9:327–336.10.2147/PPA.S74531
  20. Currie CJ, Peyrot M, Morgan CL, et al. The impact of treatment noncompliance on mortality in people with type 2 diabetes [published online April 17, 2012]. Diabetes Care 2012;35;(6):1279–1284.10.2337/dc11-1277
  21. Mazzaglia G, Ambrosioni E, Alacqua M, et al. Adherence to antihypertensive medications and cardiovascular morbidity among newly diagnosed hypertensive patients [published online October 5, 2009]. Circulation 2009;120;(16):1598–1605.10.1161/CIRCULATIONAHA.108.830299
  22. Shafrin J, Forma F, Scherer E, et al. The cost of adherence mismeasurement in serious mental illness: a claims-based analysis. Am J Manag Care 2017;23;(5):e156–e163.
  23. Gilmer TP, Dolder CR, Lacro JP, et al. Adherence to treatment with antipsychotic medication and health care costs among Medicaid beneficiaries with schizophrenia. Am J Psychiatry 2004;161;(4):692–699.10.1176/appi.ajp.161.4.692
  24. Hansen RA, Maciejewski M, Yu-Isenberg K, Farley JF. Adherence to antipsychotics and cardiometabolic medication: association with health care utilization and costs. Psychiatr Serv 2012;63;(9):920–928.10.1176/appi.ps.201100328
  25. Kessler RC, Heeringa S, Lakoma MD, et al. Individual and societal effects of mental disorders on earnings in the United States: results from the national comorbidity survey replication [published online May 7, 2008]. Am J Psychiatry 2008;165;(6):703–711.10.1176/appi.ajp.2008.08010126
  26. De Bleser L, Vincke B, Dobbels F, et al. A new electronic monitoring device to measure medication adherence: usability of the Helping Hand™ [published online March 1, 2010]. Sensors (Basel) 2010;10;(3):1535–1552.10.3390/s100301535
  27. Profit D, Rohatagi S, Zhao C, et al. Developing a digital medicine system in psychiatry: ingestion detection rate and latency period. J Clin Psychiatry 2016;77;(9):e1095–e1100.10.4088/JCP.16m10643
  28. Rohatagi S, Profit D, Hatch A, et al. Optimization of a digital medicine system in psychiatry. J Clin Psychiatry 2016;77;(9):e1101–e1107.10.4088/JCP.16m10693
  29. Aitken M, Clancy B, Nass D. The growing value of digital health: evidence and impact on human health and the healthcare system. IQVIA Institute for Human Data Science November 2017; Available at: https://regresearchnetwork.org/wp-content/uploads/the-growing-value-of-digital-health.pdf. Accessed April 24, 2019
  30. Mack H. Quio gets $1.05M for connected drug injector. Mobi-HealthNews October 3, 2016; Available at: https://www.mobihealthnews.com/content/quio-gets-105m-connected-drug-injector. Accessed April 24, 2019
  31. Karve S, Cleves MA, Helm M, et al. Prospective validation of eight different adherence measures for use with administrative claims data among patients with schizophrenia [published online April 23, 2009]. Value Health 2009;12;(6):989–995.10.1111/j.1524-4733.2009.00543.x
  32. Kamat SA, Rajagopalan K, Pethick N, Willey V, Bullano M, Hassan M. Prevalence and humanistic impact of potential misdiagnosis of bipolar disorder among patients with major depressive disorder in a commercially insured population. J Manag Care Pharm 2008;14;(7):631–642.
  33. Lee TA, Pickard AS, Bartle B, Weiss KB. Osteoarthritis: a comorbid marker for longer life?. Ann Epidemiol 2007;17;(5):380–384.10.1016/j.annepidem.2007.01.033
  34. Jena AB, Mechoulan S, Philipson TJ. Altruism and innovation in health care. J Law Econ 2010;53;(3):497–518.10.1086/648383
  35. CLOZARIL (clozapine) tablets prescribing information East Hanover, New Jersey: Novartis Pharmaceuticals Corporation. 2017; Available at: http://clozaril.com/wp-content/themes/eyesite/pi/Clozaril_PI.pdf. Accessed April 24, 2019
  36. Patel NC, Crismon ML, Miller AL, Johnsrud MT. Drug adherence: effects of decreased visit frequency on adherence to clozapine therapy. Pharmacotherapy 2005;25;(9):1242–1247.10.1592/phco.2005.25.9.1242
  37. Agency for Healthcare Research and Quality. Treatment for bipolar disorder June 23, 2014; Available at: https://effectivehealthcare.ahrq.gov/products/bipolar-disorder-treatment/research-protocol/. Accessed April 24, 2019
  38. Gartlehner G, Hansen RA, Morgan LC, et al. Second-generation antidepressants in the pharmacologic treatment of adult depression: an update of the 2007 comparative effectiveness review Agency for Healthcare Research and Quality. December 5, 2011; Available at: https://effectivehealthcare.ahrq.gov/products/antide-pressants-update/research. Accessed April 24, 2019
  39. Bennett WL, Wilson LM, Bolen S, et al. Oral diabetes medications for adults with type 2 diabetes: an update Agency for Healthcare Research and Quality. March 15, 2011; Available at: https://effectivehealthcare.ahrq.gov/products/type-2-diabetes-medicationsupdate/research. Accessed April 24, 2019
  40. James PA, Oparil S, Carter BL, et al. 2014 evidence-based guideline for the management of high blood pressure in adults: report from the panel members appointed to the eighth joint national committee (JNC 8). JAMA 2014;311;(5):507–520.10.1001/jama.2013.284427
  41. Weber MA, Schiffrin EL, White WB, et al. Clinical practice guidelines for the management of hypertension in the community: a statement by the American Society of Hypertension and the International Society of Hypertension [published online December 17, 2013]. J Clin Hypertens(Greenwich) 2014;16;(1):14–26.10.1111/jch.12237
  42. Pharmacy Quality Alliance. Adherence: PQA adherence measures August 28, 2018; Available at: https://www.pqaalliance.org/adherence-measures. Accessed April 25, 2019
  43. National Committee for Quality Assurance. Adherence to antipsychotic medications for individuals with schizophrenia (SAA) 2018; Available at: https://www.ncqa.org/hedis/measures/adherence-to-antipsychotic-medications-for-individuals-withschizophrenia/. Accessed April 25, 2019
  44. Stephenson JJ, Tunceli O, Gu T, et al. Adherence to oral second-generation antipsychotic medications in patients with schizophrenia and bipolar disorder: physicians’ perceptions of adherence vs. pharmacy claims [published online May 11, 2012]. Int J Clin Pract 2012;66;(6):565–573.10.1111/j.1742-1241.2012.02918.x
  45. Haddad PM, Brain C, Scott J. Nonadherence with antipsychotic medication in schizophrenia: challenges and management strategies. Patient Relat Outcome Meas 2014;5:43–62.10.2147/PROM.S42735
  46. Marengoni A, Onder G, Degli Esposti L, et al. Adherence to selective serotonin and serotonin-norepinephrine reuptake inhibitor prescriptions affects overall medication adherence in older persons: evidence from the Italian nationwide OsMed Health-DB Database. J Clin Psychiatry 2016;77;(12):1712–1718.10.4088/JCP.15m10503
  47. Kreyenbuhl J, Dixon LB, McCarthy JF, et al. Does adherence to medications for type 2 diabetes differ between individuals with vs without schizophrenia [published online August 20, 2008]?. Schizophr Bull 2010;36;(2):428–435.10.1093/schbul/sbn106
  48. Agency for Healthcare Research and Quality. The AHRQ Multiple Chronic Conditions Research Network December 2012; Available at: https://www.ahrq.gov/patient-safety/settings/long-term-care/resource/multichronic/index.html. Accessed April 26, 2019
  49. Agency for Healthcare Research and Quality. Defining the PCMH Patient Centered Medical Home Resource Center. Available at: https://www.pcmh.ahrq.gov/page/defining-pcmh. Accessed August 21, 2017
  50. Shafrin J, May SG, Shrestha A, et al. Access to credible information on schizophrenia patients’ medication adherence by prescribers can change their treatment strategies: evidence from an online survey of providers. Patient Prefer Adherence 2017;11:1071–1081.10.2147/PPA.S135957
  51. Ingersoll KS, Cohen J. The impact of medication regimen factors on adherence to chronic treatment: a review of literature. J Behav Med 2008;31;(3):213–224.10.1007/s10865-007-9147-y