Loeppke, Ronald MD, MPH; Haufle, Vince MPH; Jinnett, Kim PhD; Parry, Thomas PhD; Zhu, Jianping PhD; Hymel, Pamela MD, MPH; Konicki, Doris MHS
The expanding burden of chronic illness, driven by precursor health risks, is leading to a health crisis that dominates the cost crisis in health care. It has been estimated that 133 million Americans have at least one chronic health condition and that 75% of all medical expenses are related to the treatment of those with one or more chronic conditions.1,2 In fact, chronic conditions drive 96% of the costs in the Medicare system and 83% of the costs of the Medicaid system, and they are responsible for two-thirds of the rise in overall health care costs in the United States since 1980.3,4
A 2009 study of employers' health care costs demonstrated that chronic health problems—including diabetes, depression, hypertension, heart disease, and high cholesterol—were among the costliest for employers.5 When absenteeism (absence from work) and presenteeism (decreased job performance while at work) costs were included in the mix, the costliest conditions for employers were still attributed to chronic health problems.
Chronic health conditions, in many circumstances, are preventable.3,4 Therefore, clinical preventive medicine strategies are critical elements of the solution to this cost conundrum of chronic illness and its associated consequences of morbidity and mortality. Three levels of prevention are needed:
* Primary prevention strategies include health promotion, health protection, health education, lifestyle management, safety engineering, job ergonomics and organizational design, nutrition, prenatal care, immunizations, and other wellness services because they help people stay healthy, safe, and productive.
* Secondary prevention strategies include screening and early detection programs, health coaching, and biometric testing. These strategies can identify conditions earlier than they would have been by typical clinical manifestation.
* Tertiary prevention strategies include medication management/adherence and disease management, evidence based quality care management, return to work programs and disability management. These strategies can provide earlier interventions, limit the destructive and often disabling impact of serious medical conditions on function in daily life and work, can protect or restore productive lifestyles, and can reduce future costs.6
A study at the Milken Institute has calculated that seven chronic conditions (cancer, heart disease, hypertension, mental disorders, diabetes, pulmonary conditions, and stroke) are costing the US economy alone more than $1 trillion per year—with anticipated growth rates of the prevalence of those seven conditions to yield an illness burden of $4 trillion per year by 2023.7 However plausible estimates of potential gains (avoided losses) associated with reasonable improvements in prevention, detection, and treatment of just these seven conditions show that annual treatment costs could be cut in the US by $217 billion and reduce health-related productivity losses by $905 billion by 2023. It has been estimated that lowering obesity rates alone could lead to productivity gains of $254 billion and avoidance of $60 billion in treatment expenditures.
Taking medication as prescribed can be effective in controlling chronic health conditions and reducing the risks of complications.8–10 However, compliance with medication regimes for major chronic health conditions such as coronary artery disease,11 hypertension,12 and depression 13 is not optimal. A 2003 report by the World Health Organization estimated that worldwide only 50% of the patients with chronic diseases take their medication as prescribed.14 It is estimated that between 30% and 50% of the individuals with hypertension do not take their medications as prescribed12 and employees with hypertension have twice the illness and absenteeism rate of those without hypertension.15 Employers spend an estimated $392 per employee annually for the direct and indirect costs associated with hypertension,16 and the overall cost for hypertension in 2008 was an estimated $64.9 billion.17
Likewise, depression, affecting 35 million individuals annually18 and one of the top two chronic health conditions affecting employees,5 has a poor history of adherence to medication protocols.19 A 2007 study by Burton et al concluded that had the 2112 antidepressant users in the study been medication compliant, $400,000 in lost short term disability (STD) workdays could have been saved. 13 Birnbaum et al20 “showed that increased compliance with antidepressants is significantly associated with reduced absenteeism but not with medical cost/utilization. Compliance with antidepressant treatment appears to decrease the debilitating effects of depression, allowing people to get back to work, but because the recovery process can continue at work, it may hinder productivity.”
Noncompliance can be attributed to a number of factors, including personal characteristics, medication regimen characteristics, disease-specific factors, patient/provider and support relationships, and psychosocial factors. A number of studies have examined and found an association between these factors and medication adherence. Findings show that people who take more drugs and more doses per day have lower levels of compliance21–23 as do those who have a low literacy rate,24 depression,12,23 poor communication from the health provider,25 and minimal support from others.
A 2007 to 2009 multiemployer research study exploring methodological refinements in measuring health-related lost productivity and assessing the business implications of a full-cost approach to managing health, demonstrated that health-related workforce absence and job performance (absenteeism and presenteeism) costs are significantly greater than medical and pharmacy costs alone, on average 2.3 to 1.5 Chronic conditions such as depression/anxiety, obesity, arthritis, and back/neck pain were identified as especially important causes of productivity loss. Another finding of the study was the need to further investigate the impact of medication adherence on lost productivity costs.
Much is known regarding medication adherence for various chronic diseases, but the scientific literature has little to say about the relationship between medication adherence, comorbidity and health risk factors in the workplace—and the overall impact of all three on absence and job performance.
To address these new questions, a follow-up study was initiated in 2010 to:
1. Understand the patterns of medication usage and adherence related to comorbidity and health risk factors—particularly among subjects with several chronic conditions: depression, type 2 diabetes, hypertension, and coronary artery disease.
2. Evaluate whether medication adherence, comorbidity, and health risk factors among subjects with chronic conditions is associated with absenteeism and job performance.
3. Assess the impact of depression as a comorbidity on medication adherence, absenteeism, and job performance.
MATERIALS AND METHODS
In 2010, a retrospective observational study using employees' medical and pharmacy administrative claims data and self-reported health risk appraisal data was conducted. The study covered a period of 29 months in length, beginning January 1, 2007 and ending May 31, 2009. Employee medication adherence behavior was measured for a minimum of 183 days and a maximum of 365 days preceding the completion of a health risk appraisal survey.
Employees' medical and pharmacy claims data were used for this analysis. Claims data included health plan eligibility, inpatient and outpatient diagnoses, as well as pharmacy records. Paid amounts, as well as dates of service, were available for all claims. The health risk appraisal instrument used in this study, Alere's Health and Productivity Assessment (HPA), contains a standard set of medical and lifestyle indicators and items from the HPQ-Select survey for capturing self-reported data on health conditions, absence, and job performance. The HPQ-Select survey is an updated, employer-focused version of the Work Performance Questionnaire (HPQ)26 that was initially developed by the Dr Ronald Kessler of Harvard Medical School.
The study population included employees from five companies (Companies D, E, F, H, and J) that previously participated in the research project “Health and Productivity as a Business Strategy Study.”5 Employees in the studied companies completed 115,961 HPA surveys between January 1, 2007 and May 31, 2009 and were considered for inclusion in the study if they met the following criteria:
* Responded to all HPQ-Select survey questions within each HPA,
* Health-plan eligible for at least 11 of the 12 months prior to the HPA completion date,
* Between the ages of 18 to 64 at the time of HPA survey completion.
If an individual completed multiple HPA surveys within the study period, their first survey within the study period was selected for study inclusion. The study examined employees' last HPA within the survey period and found no substantial differences in the outcomes using the first or the last HPA.
Sample Selection by Health Condition and Medication Treatment Group
Medication adherence was reviewed among individuals with at least one of the following four chronic health conditions: coronary artery disease (CAD), hypertension, type 2 diabetes and depression. Specific ICD-9-CM (International Classification of Diseases, 9th Revision–Clinical Modification) codes used for condition selection are outlined in Appendix 1 (see supplemental appendix at http://links.lww.com/JOM/A54). Depression was also identified as a comorbidity covariate in this study. Depression—particularly in combination with other chronic health conditions—can have significant effects on sick days and health and disability costs.27
Medical and pharmacy claims data and self-reported data from the HPA survey were used to determine if an employee had been diagnosed with at least one chronic health condition of interest and to identify the existence of other chronic health conditions as a measure of comorbidity. The medical and pharmacy claims-based identification criteria required an employee to have at least one physician visit, emergency department visit or inpatient stay with a condition-related ICD-9-CM diagnosis and at least one condition-related pharmacy prescription fill within a maximum of 18 months preceding the HPA completion date. Diagnosis codes were examined up to five positions within a claim when available. National Drug Codes (NDCs) within pharmacy claims files were mapped to each chronic health condition following the methods previously reported in Loeppke et al.5
If an employee did not have a condition-related physician visit, emergency department visit or inpatient stay, a condition self-report and pharmacy claims-based identification criteria were then applied. An employee must have self-reported the chronic health condition of interest on the HPA and had at least one condition-related pharmacy fill within a maximum of 18 months preceding the HPA completion date. Both the medical claims-based and self-report-based condition identification criteria required the presence of a condition related pharmacy claim to qualify for study inclusion.
A detailed description of the ICD-9-CM diagnosis codes, HPA survey questions and condition-related drug classes used for the identification process can be found in Appendices 1 and 2 (see supplemental appendices at http://links.lww.com/JOM/A54; http://links.lww.com/JOM/A57).
Among the employees identified with at least one of the four chronic health conditions of interest, the study examined employee's medication adherence behavior for a set of nine different condition-related medication groups. Employees qualified for inclusion in a condition/medication group if they had an index prescription fill-date occurring within the 183 to 365 days preceding their HPA completion date and the prescription National Drug Code was determined to belong to the medication group.
Figure 1 provides an overview of the sample selection procedure we used to create the nine medication subsamples for the analysis. A total of 115,991 HPA surveys were completed during the study period. Health and Productivity Assessment surveys were restricted to employees only (N = 111,038) and further restricted to surveys with completed responses to the HPQ-select survey items (N = 110,904). These surveys include a total of 64,422 unique employees. Among these employees four medication subsamples were created for CAD: statin (N = 666), beta blockers (N = 478), ACE or ARB (N = 543), and antiplatelet or anticoagulant (N = 303). One subsample was created for hypertension: antihypertensive (N = 5459). Three subsamples for type 2 diabetes: statin (N = 774), ACE or ARB (N = 790), and insulin or oral hypoglycemic or metformin (N = 1312). One subsample was created for depression: antidepressant (N = 2120). A more detailed description of the full sample selection process can be found in Appendices 3 and 4 (see supplemental appendices at http://links.lww.com/JOM/A58; http://links.lww.com/JOM/A59).
This study examined one self-reported measure of absenteeism and one self-reported measure of job performance. The first measure, absenteeism, is based on employees' responses to the following two questions: “How many hours does your employer expect you to work in a typical seven-day week?” and “About how many hours altogether did you work in the past four weeks (28 days)?” The hours expected in a 7-day week is multiplied by four to calculate the number of hours expected to work in 28 days. Actual hours worked (28 days) was subtracted from the hours expected to work (28 days) to determine absenteeism hours. This measure also allowed for negative absenteeism hours, which we interpreted as working more hours than expected or working overtime.
The second measure, job performance, is based on employees' responses to the following question: “On a scale from 0 to 10 where 0 is the worst job performance anyone could have at your job and 10 is the performance of a top worker, how would you rate your overall performance on the days you worked during the past four weeks (28 days)?” Responses were multiplied by 10 to convert an employee's performance into a job performance percentage score. A high value on the job performance score indicates higher job performance.
Medication Possession Ratio
Medication Possession Ratio (MPR) was used to measure employees' compliance with medication treatment. The MPR measurement period was defined as the 365 days preceding the HPA completion date. The MPR denominator is the duration (days) from the first (index) prescription fill date within the measurement period through the last day of the measurement period (HPA completion date). The MPR numerator is the days supplied over the same period adjusting for prescription overlap. In instances where prescription days for two or more National Drug Codes overlapped, only 1 day of possession was counted for each day of overlap. Several inclusion and exclusion criteria were also applied to the MPR measure. The index prescription fill-date must have occurred within the 183 to 365 days preceding the HPA completion date. The study excluded any “carry in” days for prescriptions filled before the beginning of the measurement period and excluded “carry out” days when medication supply goes beyond the end of the measurement period. Each employee's MPR numerator/denominator set was limited such that the numerator is less than or equal to the denominator (ie, Days Supplied <= Denominator Days; an employee's MPR cannot exceed 100%). The study only included prescriptions from predefined condition-specific drug classes. Two different constructions of the MPR were examined. The first was a continuous version of the variable and the second was a dichotomous version of the variable (MPR ≥ 0.8) used to indicate a threshold of adherence.28
A hybrid comorbidity index was used for this study based on the combination of self-reported responses to the HPQ-Select's chronic health condition assessment within the HPA and a modified version of the ICD-9-CM-based Charlson comorbidity index.29 Condition identification via ICD-9-CM codes was based on the medical claims incurred during the 12 months preceding the HPA survey date. A total of 31 chronic health conditions were assessed with the claims-based, self-report-based, or both approach. A list of the conditions and the assessment sources can be found in Appendix 5 (see supplemental appendix at http://links.lww.com/JOM/A55).
Health Risk Status
Following the approach of Edington and colleagues,30 an index of overall health risk—low (0 to 2 risks), moderate (3 to 4 risks), and high (5 or more risks)—was created based on the risk status of 12 modifiable medical and lifestyle risk indictors. These include: body weight, blood pressure, cholesterol, triglycerides, blood glucose, tobacco use, alcohol use, physical activity, dietary fat, fruits and vegetable consumption, stress/coping, and seat belt use (see supplemental Appendix 6 for at-risk thresholds at http://links.lww.com/JOM/A56). These indicators, individually and collectively, have been shown to relate to preventable chronic health conditions, and health care and productivity costs.30,31
Average Monthly Paid Health Care Costs
Employee's health care costs were defined as the average monthly paid costs for all physician visits, emergency department visits, inpatient stays, and prescription drugs over the 12 months preceding the HPA completion date. This includes employees' total health care costs—not condition-specific costs alone.
A claim-based depression, self-report-based depression, or both identification criteria were used to identify the presence of depression comorbidity among employees with CAD, hypertension, or type 2 diabetes.
Additional covariates included age at HPA completion, gender, marital status, race/ethnicity, and job-type classification.
The primary hypothesis tested was whether individuals with CAD, hypertension, type 2 diabetes, or depression who are adherent to medication treatment with a known therapeutic use (for CAD, hypertension, type 2 diabetes, or depression) will report less absence and higher job performance. The secondary hypothesis tested was whether individuals with a chronic condition and depression comorbidity would report more absence and lower job performance.
Descriptive statistics (ie, means, standard deviations, percentages of total) of all model variables were calculated for each medication subsample. Ordinary least squares regression was used to examine the relationship between medication adherence (MPR) and the two outcome measures (absence and job performance) for each of the nine medication subsamples (controlling for all covariates).
Table 1 displays the descriptive statistics for all model variables across the nine medication subsamples. Sample sizes range between n = 303 (CAD: antiplatelet or anticoagulant) and n = 5459 (hypertension: antihypertensive).
TABLE 1-a. Descripti...Image Tools
TABLE 1-b. Descripti...Image Tools
The mean age across all medication subsamples varies from 45.0 years of age (depression: antidepressant) to 55.1 years of age (CAD: antiplatelet or anticoagulant). Gender ranges between 44.2% male (depression: antidepressants) to 86.2% male (CAD: statin). Across all medication subsamples whites are the predominant ethnic group—58.2% (type 2 diabetes: insulin or oral hypoglycemic or metformin) to 83.6% (depression: antidepressant). Asians represent between 18.7% and 22.8% of the type 2 diabetes medication subsamples. All other ethnic groups represent 15% or less of the sample across all medication subsamples. Between 72.5% and 76.7% of employees across all subsamples are married or in domestic partnerships with the exception of those in the depression: antidepressant group with 59.5% married/domestic partnership.
Between 29.5% and 39.7% of all subsamples are in the professional job classification. Between 10.1% and 14.7% of all subsamples are in the executive, administrator, or senior manager group. Service occupations vary between 8.4% (depression: antidepressant) and 18.2% (CAD: antiplatelet or anticoagulant). About 10% of the CAD: statins and CAD: antiplatelet or anticoagulant are in the clerical and administrative support classification, with the highest percent of this job classification in the depression: antidepressant group (17.4%). All other job classifications represent 10% or less of the subsamples across all groups, except operator or laborers in the CAD: antiplatelet or anticoagulant group at 11.2% of that subsample.
Comorbidity Index, Average Paid Claims
Across all subsamples employees have between 4.2 and 5.8 comorbidities with average monthly paid claims cost (all medical and pharmacy costs combined) for the 12 months preceding the HPA ranging between $509 for hypertension: antihypertensive and $1572 for CAD: antiplatelet or anticoagulant.
Health Risk Status
Most employees across all subsamples tend to be in the medium health risk group (between 41.2% and 45.9%) with about a third or more in the highest risk group (29.4%–37.8%) and fewer in the low risk group (16.7%–24.8%) than in typical employed populations.
By definition, all (100%) of the employees in the depression: antidepressant group have a depression diagnosis. For all other subsamples, depression comorbidity ranges between 8.3% (type 2 diabetes: insulin or oral hypoglycemic or metformin) and 11.7% (CAD: beta-blocker).
Absenteeism and Job Performance
Mean absenteeism over 28 days varies between 2.4 and 8.3 hours. Mean job performance varies between 81.6% and 87.0% present while at work over 28 days.
Medication Possession Ratio
Table 2 displays the MPR across all nine-medication subsamples. The average MPR varies between 0.70 (depression: antidepressant) to 0.79 (CAD: beta blockers, hypertension: antihypertensives and diabetes 2: ACE inhibitors or ARB). The percent with an MPR of at least 0.80 varies between 53.0% (diabetes 2: statin) and 65.6% (hypertension: antihypertensives).
Table 3 is a summary of the major findings from the 36 full model regression analyses across all nine-medication subsamples. Except for the continuous and categorical beta effects, the covariate beta effects presented in Table 3 are from the continuous MPR models only. R-squared values for the nine absenteeism medication subsample models ranged from 0.04 (type 2 diabetes: ACE or ARB) to 0.11 (CAD: antiplatelet or anticoagulant). R-squared values for the nine job performance medication subsample models ranged from 0.06 (depression: antidepressant) to 0.15 (CAD: antiplatelet or anticoagulant). A detailed summary of all model results can be found in Appendices 7 and 8 (see supplemental appendices at http://links.lww.com/JOM/A60; http://links.lww.com/JOM/A61).
Medication Adherence—Continuous and Categorical MPR
Medication adherence (both continuous and categorical MPR) is a significant (P < 0.05) predictor of absenteeism in the direction hypothesized for one (CAD: statin) of the nine medication subsamples. Among the continuous MPR model, a one percentage point increase in MPR for CAD: statin translates into 0.217 fewer absence hours missed over a 28-day period (eg, moving from an MPR of 60% to 61% leads to an increase in approximately 13 minutes of less-absence over a 28-day period). Among the categorical MPR model, those with a CAD: statin MPR greater than or equal to 80% had 9.02 fewer hours absent from work over a 28-day period (1.28 days less absence per month) compared to those with a CAD: statin MPR of less than 80%.
Medication adherence (categorical MPR) is a significant predictor of job performance in the direction hypothesized for one (type 2 diabetes: insulin or oral hypoglycemic or metformin) of the nine medication subsamples. On average, individuals with an MPR greater than or equal to 80% are 1.46 percentage points higher on the job performance scale than individuals with MPRs that are lower than 80% for type 2 diabetes: insulin or oral hypoglycemic or metformin. In other words, those with an MPR greater than or equal to 80% have 2.34 hours more work performance over a 28-day period than those with an MPR less than 80%.
Health Condition Comorbidity
The number of health condition comorbidities was found to be a significant predictor of absenteeism in five of the nine medication subsamples. The effect of comorbidity ranged between 1.4 and 3.0 absenteeism hours over a 28-day period for each additional comorbidity. Alternatively, a significant relationship between the number of health condition comorbidities and job performance was only seen in one of the nine subsamples. Among individuals with type 2 diabetes insulin or oral hypoglycemic or metformin, for each comorbid condition, the job performance percentage declines −0.53 percentage points.
There were no significant absence effects among individuals with depression comorbidity across the eight subsamples. However, a significant relationship between depression comorbidity and job performance was found among five of the eight subsamples. In these five cases having depression comorbidity was associated with between a 4.0 and 5.5 percentage point decrease in job performance.
Health Risk Status
High or moderate health risk status was not found to be a significant predictor of absenteeism across any of the nine subsamples. In contrast, high-risk status was a significant predictor of lower job performance at work in all nine subsamples and moderate risk status was a significant predictor of job performance in two of the nine subsamples. For example, the high-risk effect varied between 3.6 and 5.7 lower job performance percentage points compared to those in the low risk group.
In this extension of the work begun in the 2007–2009 “Health and Productivity as a Business Strategy” study, significant new findings help to bolster the case for integrated health management strategies in the workplace.
While statistically significant conclusions about medication adherence impacting worker absence and job performance across all conditions could not be drawn from the largely medication-compliant cohort in this study, the complexity and impact of chronic comorbidities and health risks on absence and job performance, respectively, was clear and substantial. Specifically, new evidence was gathered suggesting strong links between the number of chronic comorbidities and absenteeism in the workplace and between health risks and job performance.
Across all subsamples in the study, employees have between 4.2 and 5.8 comorbidities. Most employees across all subsamples tend to be in the medium health risk group (between 41.2% and 45.9%), with about a third or more in the highest risk group (29.4% to 37.8%) and fewer in the low risk group (16.7% to 24.8%) than in typical employed populations. The mean absenteeism over 28 days varies between 2.4 and 8.3 hours for the study group while the mean job performance varies between 81.6% and 87.0% present while at work over 28 days.
The average MPR for the study group varies between 0.70 (depression: antidepressant) to 0.79 (CAD: beta blockers, hypertension: antihypertensives, and diabetes 2: ACE inhibitors or ARB). The percent with an MPR of at least 0.80 varied between 53.0% (diabetes 2: statin) and 65.6% (hypertension: antihypertensives).
When we compared mean medication adherence levels among the various medication subsamples to what has been published in the literature, we conclude that our study population has comparable to slightly higher overall medication adherence on average.32,33
Even though this study population may not have allowed enough adherence variation to detect significant impact on absence and job performance across several conditions and medications, it is interesting that among the continuous MPR model, for each 10% increase in MPR for CAD: statin, that figure translated into 2.17 fewer absence hours missed over a 28-day period. We also found that job performance was 1.5 points higher for those with MPRs greater than or equal to 80% for type 2 diabetes: insulin or oral hypoglycemic or metformin.
One of the confounding variables that could not be ascertained in this study, and which could be influencing inconsistent results across conditions and adherence levels, is that even though individuals are apparently highly adherent to taking their medications at the dosage their doctor has prescribed, it is not known whether the patients are receiving the right dosage to bring them into optimal therapeutic ranges and whether their conditions are being well managed.
Also, as the number of health condition comorbidities was found to be a significant predictor of absenteeism in five of the nine medication subsamples, it appears that multiple comorbidities are a more consistent predictor of work absence than level of medication adherence. By contrast, when patients had depression as one of the comorbid conditions, it showed a significant relationship on job performance rather than absenteeism. Further study is needed to determine if this effect might be related to whether people with multiple comorbidities were more apt to show up to work when their depression was diagnosed and treated even though the side-effects of the antidepressant or interaction with other medications led to lower performance while at work.
In the original Health and Productivity as a Business Strategy study, employees with one comorbid condition comprised nearly 40% of comorbid cases and generated about 15,600 lost days from absence and lower job performance. Employees with six or more comorbidities made up about 8% of the group but contributed more than 36,000 lost days.5 This study shows in more granular detail the increasing complexity of the treatment picture when comorbidities are involved.
One of the more compelling findings from this study is that high health risk status was a significant predictor of lower job performance in all nine subsamples and moderate risk status was a significant predictor of lower job performance in two of the nine subsamples. This relationship between higher health risk and lower job performance has been shown previously.30 However, it is enlightening to find evidence that higher health risks impact performance at work—even in a population of individuals with chronic conditions that are highly compliant with their medications.
Clearly, the complexity and impact of comorbidities and health risks must be recognized as important components in a truly holistic approach to health management. This underscores the premise that when providing personal care management, as well as population health management, it is as important to understand the type of person (each with a unique constellation of health risks and comorbidities) as it is to understand the type of medical condition the person has. It is important to deal with the whole person—not only their medical issues, but also the emotional, mental, physical, and social aspects of their well-being.
More detailed study is needed on the connection between health risk status, comorbidity, and medication adherence and impact on the full costs of illness, including health-related productivity loss.
However, statistically significant conclusions can be drawn from this study regarding impacts found in the relationship between health risk factors, comorbidity and impact on absence and job performance in a population with high prevalence of chronic illness, yet high medication-adherence.
First, the study provides more granular corroborating evidence of the impact of comorbidity on absenteeism levels—that is, greater the level of comorbidity, the greater the impact on absence from work (P < 0.05).
Second, the study suggests significant impacts from health risk factors on treatment efficacy for chronic disease. In the highly medication-adherent sample population studied, high health risk status was a significant predictor of lower job performance in all of the targeted subsamples (P value range < 0.05 to < 0.0001).
Other significant findings included:
* Statin medication adherence in individuals with CAD was a significant predictor (P < 0.05) of decreasing absenteeism from work.
* Among the continuous MPR model, a one-percentage point increase in MPR for CAD: statin translates into 0.217 fewer absence hours missed over a 28-day period (eg, moving from an MPR of 50%–70% leads to an increase in approximately 4.33 hours of less-absence over a 28-day period).
* Among the categorical MPR model, those with a statin MPR greater than or equal to 80% had 9.02 fewer hours absent from work over a 28-day period (1.28 days less absence per month) than those with an MPR less than 80%.
* Insulin or oral hypoglycemic or metformin medication adherence in type 2 diabetics was a significant (P = 0.038) predictor of improved job performance.
* Those individuals with an MPR equal to or greater than 80% had 2.34 hours more work performance over a 28-day period than those with an MPR less than 80%.
* Among individuals with type 2 diabetes on insulin or oral hypoglycemic or metformin, for each comorbid condition, their work performance declined −0.53%.
* By contrast, when patients had depression as one of the comorbid conditions, it showed a significant relationship on lower job performance rather than absenteeism (P < 0.0001).
The fact that health risk status remained a significant predictor of reduced performance while at work even in a highly medication-adherent population suggests that while medication adherence is a critical element of tertiary prevention initiatives in conjunction with effective disease management in reducing the acuity and disability impact of the illness, it is also important to reduce the health risks and increase the well-being of the person.
These results suggest that employer-driven integrated health and productivity management strategies should include an emphasis on primary and secondary prevention to reduce health risks even in the segment of the population that has chronic conditions and is already receiving tertiary prevention services through disease management and medication management. Research has shown that chronic health conditions are largely preventable3,4 and several studies have shown that health risks can be reduced with appropriate interventions.30,34 Therefore, integrated clinical preventive medicine strategies should be part of the foundational underpinnings of the solution to the cost conundrum of this economic burden of chronic illness and its associated consequences of morbidity and mortality.
The authors thank Kimberly Westrich, Bobby DuBois, MD, Gary Persinger, and Jeffrey Warren, MPA, of the National Pharmaceutical Council as well as Paul Larson of ACOEM for their assistance in providing review and comment during the study and the development of this article.
1. Thorpe KE. The rise in health care spending and what to do about it. Health Aff. 2005;24:1436–1445.
2. Thorpe KE, Florence CS, Howard DH, Joski P. The impact of obesity on rising medical spending. Health Aff. 2004;23:480–486.
3. Thorpe K. Factors accounting for the rise in health-care spending in the United States: the role of rising disease prevalence and treatment intensity. Public Health. 2006;120:1002–1007.
4. Kenneth T Keynote. Thorpe K. Presentation at: American College of Preventive Medicine 2008 Conference; February 20–23, 2008; Austin, Texas.
5. Loeppke R, Taitel M, Haufle V, Parry T, Kessler RC, Jinnett K Health and productivity as a business strategy: a multiemployer study. J Occup Environ Med. 2009;51:411–423.
6. Loeppke R The value of health and the power of prevention. Int J Workplace Health Manage. 2008;1:95–108.
7. Devol R, Bedroussian A. An Unhealthy America: The Economic Burden of Chronic Disease—Charting a New Course to Save Lives and Increase Productivity and Economic Growth. Santa Monica, Calif: Milken Institute; 2007
8. Mojtabai R, Olfson M Medication costs, adherence, and health outcomes among medicare beneficiaries. Health Aff. 2003;22:220–229.
9. Ho P, Magid D, Shetterly S, et al. Medication nonadherence is associated with a broad range of adverse outcomes in patients with coronary artery disease. Am Heart J. 2008;155:772–779.
10. Denys TL, Nau DP Oral antihyperglycemic medication nonadherence and subsequent hospitalization among individuals with type 2 diabetes. Diabetes Care. 2004;27:2149–2153.
11. Maddox TM, Ho PM Medication adherence and the patient with coronary artery disease: challenges for the practitioner. Curr Opin Cardiol. 2009;24:468–472.
12. Wang PS, Bohn RL, Knight E, Glynn RJ, Mogun H, Avorn J Noncompliance with antihypertensive medications. The impact of depressive symptoms and psychosocial factors. J Gen Int Med. 2002;17:504–511.
13. Burton WN, Chen CY, Conti DJ, Schultz AB, Edington DW The association of antidepressant medication adherence with employee disability absence. Am J Manag Care. 2007;13:105–112.
15. Tsai SP, Wendt JK, Ahmed FS, Donnelly RP, Strawmyer TR Illness absence patterns among employees in a petrochemical facility: impact of selected health risk factors. J Occup Environ Med. 2005;47:838–846.
16. Burton WN, Pransky G, Conti DJ, Chen CY, Edington DW The association of medical conditions and presenteeism. J Occup Environ Med. 2004;46(suppl 6):38–45.
17. Rosamond W, Flegal K, Furie K, et al. Heart disease and stroke statistics- update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation. 2008;117:e25–e146.
18. Kessler RC, Berglund P, Demler O, et al. National Comorbidity Survey Replication. The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R). JAMA. 2003;289:3095–3105.
19. Keller MB, Hirschfeld RM, Demyttenaere K, Baldwin DS Optimizing outcomes in depression: focus on antidepressant compliance. Int Clin Psychopharmacol. 2002;17:265–271.
20. Birnbaum HG, Ben-Hamadi R, Kelley D, et al. Assessing the relationship between compliance with antidepressant therapy and employer costs among employees in the United States. J Occup Environ Med. 2010;52:115–124.
21. Chapman RH, Benner JS, Petrilla AA, et al. Predictors of adherence with antihypertensive and lipid-lowering therapy. Arch Int Med. 2005;165:1147–1152.
22. Krousel-Wood M, Thomas S, Muntner P, Morisky D Medication adherence: a key factor in achieving blood pressure control and good clinical outcomes in hypertensive patients. Curr Opin Cardiol. 2004;19:357–362.
23. Siegel D, Lopez J, Meier J Antihypertensive medication adherence in the Department of Veterans Affairs. Am J Med. 2007;120:26–32.
24. Nichols-English G, Poirer S Optimizing adherence to pharmaceutical care plans. J Am Pharma Assc. 2000;40:475–485.
25. Harmon G, Lefante J, Krousel-Wood M Overcoming barriers: the role of providers in improving patient adherence to antihypertensive medications. Curr Opin Cardiol. 2006;21:310–315.
26. Kessler RC, Barber C, Beck A, et al. The World Health Organization Health and Work Performance Questionnaire (HPQ). J Occup Environ Med. 2003;45:156–174.
27. Katon W The impact of depression on workplace functioning and disability costs. Am J Manag Care. 2009;15:S322–S327.
28. Sikka R, Xia F, Aubert RE Estimating medication persistency using administrative claims data. Am J Manag Care. 2005;11:449–457.
29. 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.
30. Edington DW. Zero Trends – Health as a Serious Economic Strategy. Ann Arbor, MI: Health Management Research Center; 2009.
31. Goetzel RZ, Anderson DR, Whitmer RW, Ozminkowski RJ, Dunn RL, Wasserman J Health Enhancement Research Organization (HERO) Research Committee. The relationship between modifiable health risks and health care expenditures: an analysis of the multi-employer HERO health risk and cost database. J Occup Environ Med. 1998;40:843–854.
32. Yeaw J, Benner JS, Walt JG, Slan S, Smith DB Comparing adherence and persistence across 6 chronic medication classes. J Manag Care Pharm. 2009;15:728–740.
33. Cramer JA, Benedict A, Muszbek N, Keskinasian A, Khan ZM The significance of compliance and persistence in the treatment of diabetes, hypertension and dyslipidaemia: a review. Int J Clin Pract. 2008;62:76–87.
34. Loeppke R, Edington D, Beg S Impact of The Prevention Plan on Employee Health Risk Reduction. Popul Health Manag. 2010;13:275–284.