Sex differences in healthcare expenditures among adults with diabetes: evidence from the medical expenditure panel survey, 2002–2011

Diabetes is the seventh leading cause of death, accounting for significant morbidity and mortality and 20% of total health care dollars in the United States [1]. The American Diabetes Association (ADA) approximates total costs for diabetes care in the United States to be around $245 billion secondary to high indirect and direct medical costs, lost productivity, and increased disability [2, 3]. In the United States, individuals with diabetes have average medical expenditures of nearly $8000 for diabetes care alone, a cost 2.3 times higher than costs for individuals without diabetes [2‐4]. Recent evidence estimates total direct expenditures for individuals with diabetes in the United States to be about $218.6 billion unadjusted and $46 billion adjusted per year [4]. When diagnosed earlier in life (i.e., 40 years old), the excess lifetime medical spending for people with diabetes in the United States has been estimated at $124,600 per person when discounted (and approximately $211,400 per person without any discounts) compared to individuals without diabetes [1]. Ultimately, as the prevalence of diabetes in America continues to increase and the population continues to age, overall healthcare expenditures, including out-of-pocket (OOP) costs, associated with diabetes management will continue to rise, unless prevention is emphasized and the risk for complications reduced [1, 2, 5].

Evidence shows that, from the time of diagnosis, expenditures associated with diabetes care occur as a result of inpatient hospitalizations (43%), prescription medications for comorbid conditions and diabetes-related complications (18%), diabetes medications and supplies (12%), physician office-based visits (9%), and stays at skilled nursing and residential care facilities (8%) [2, 6, 7]. Between 1993 and 2006 alone, more than a 65% increase in discharges from inpatient hospitalizations were reported for patients with diabetes [8]. According to data from the National Ambulatory Medical Care Survey (NAMCS) and National Hospital Ambulatory Medical Care Survey (NHAMCS), adult patients (≥18 years of age) with diabetes made approximately 291,922,000 visits to the offices of primary care providers or hospital outpatient clinics for care between 2006 and 2010 [9]. Given the intricate nature of diabetes management, which often results in higher rates of healthcare utilization and services, it is no surprise that costs associated with diabetes care are high for adults with diabetes and continue to escalate.

For optimal diabetes management, patients with diabetes need comprehensive care, which can involve multiple and various types of healthcare services, resulting in increased OOP and overall expenses. This is especially true for non-Hispanic Blacks and women, who have been shown to have higher total health care expenses for diabetes compared to other racial and ethnic groups and men, respectively [1‐3]. Generally, women tend to have significantly higher healthcare expenditures and use more healthcare services than men [10‐12]. This is believed to be due, in part, to a higher rate of chronic conditions such as diabetes, hypertension, and obesity, in women compared to men [6, 13]. Women, compared to men, often report multiple comorbid chronic conditions [14], which leads to more prescribed medications, increased numbers of provider visits, and ultimately, higher healthcare costs [6, 13‐15]. Evidence suggests this trend remains true between women [15] and men diagnosed with diabetes [1]; however, the evidence is sparse. Therefore, given little is known about these differences and the need for additional evidence, the aim of this study was to assess sex differences in OOP and total healthcare expenditures among adults with diabetes, with the hypothesis that women with diabetes will have higher expenditures compared to men with diabetes.

Table 1 shows a summary of the sample demographics for adults with diabetes (≥18 years of age) in this study. The unweighted sample size was 20,442 for adults with diabetes, which represented a weighted sample of 17,820,243 within the U.S. population. Of the unweighted sample, 44% identified as male and 56% as female. Approximately 47% of the sample was between the ages of 45 and 64, while 40% was 65 years of age and older. Sixty-four percent of the sample was NHW, followed by 15% NHB, 14% Hispanics, and 7% of other racial and ethnic backgrounds. Fifty-nine percent was married, and 74% had a high school diploma or higher. Sixty-one percent had private insurance, and the majority (80%) lived in an urban area (i.e., MSA). Forty percent of the sample was from the South, with approximately 20% representing the Midwest and West regions. Approximately one-third of the sample rated their income as either middle or high as associated with the federal poverty line. Finally, statistically significant differences in age, race/ethnicity, marital status, educational level, insurance coverage, income, and comorbid conditions including cardiovascular disease, joint pain, arthritis, and asthma were observed between men and women in the sample.

Table 2 shows the unadjusted mean of health services for out-of-pocket expenses and total direct expenditures by sex among adults with diabetes. The OOP costs for the pooled sample was statistically different with women spending $1878 compared to men spending $1631 (95% Confidence Interval (CI) $1788–$1969 vs. $1544–$1717, respectively; p < 0.001). For specific health services, women were found to have higher OOP expenses and total direct expenditures compared to men. Statistically significant differences were observed in OOP expenses for prescriptions with women having approximately $218 more OOP expenses in a given year compared to men ($959; 95% CI $918–$1000 vs. $1177; 95% CI $1117–$1237; p < 0.001). Similarly, statistical differences in prescription costs were observed between men and women for total direct expenditures. In a given year, prescription costs for women totaled approximately $3797 (95% CI $3660–$3934) compared to $3334 (95% CI $3208–$3460) for men (p < 0.001). In addition, women had higher total direct expenditures within a year for home healthcare compared to men ($752; 95% CI $646–$858 vs. $397; 95% CI $332–$462; p < 0.001).

Table 3 shows the adjusted incremental effects of healthcare expenditures by sex among adults with diabetes. For OOP expenses, women had higher costs within a given year for prescriptions ($156; 95% CI $87–$225; p < 0.001), office-based visits ($53; 95% CI $24–$83; p < 0.001), and other healthcare expenses ($19; $7-$30; p < 0.001) compared to men. The pooled sample paid approximately $242 (95% CI $134–$350; p < 0.001) OOP for healthcare services. Similar to observations for OOP expenses, women paid approximately $184 (95% CI $50–$318; p = 0.007) in total direct expenditures for prescriptions. They also spent more in total direct expenditures for home health services ($59; 95% CI $2–$116; p = 0.041). A statistically significant difference in total direct expenditures was not observed in the pooled sample; however, it was found among the pooled sample for OOP expenses, which suggests different factors may be driving OOP expenses in women compared to men that are causing them to have higher costs.

In this sample of adults with diabetes, women with diabetes had statistically higher healthcare expenditures compared to men, after controlling for relevant confounding factors. In unadjusted analyses, women had significantly higher mean OOP and total expenditures for prescription services compared to men. In addition, women had higher total direct expenditures for home healthcare in comparison to men. After adjusting for sociodemographic characteristics, comorbidities, and time, these statistically significant differences persisted. In adjusted analyses, women had higher incremental OOP costs for prescription medications, office-based visits, and other services, and also had significantly higher total direct expenditures for prescriptions and home healthcare services compared to men with diabetes. While higher OOP and total direct expenditures for women might reflect higher utilization, it is imperative to determine best practices for keeping costs low and reducing the financial burden of diabetes care for women.

Prior evidence has demonstrated higher healthcare expenditures in women with diabetes. In this sample, women with diabetes were found to have significantly higher expenditures compared to men, particularly for healthcare services including office-based visits, home healthcare, and prescriptions. Zhou et al. found women with diabetes to have higher estimated annual medical spending and excess lifetime incremental expenses compared to men with diabetes [1]. Similarly, in a cohort of adults with diabetes in the United States, followed from 1997 to 2006, women had higher healthcare expenditures than men [32]. For example, NHW women with a BMI > 40 kg/m had lifetime healthcare expenditures in the amount of $185,609, and NHW women with a BMI between 18.5 and 24.9 kg/m had lifetime healthcare expenditures that totaled $183,704 compared to men [32]. More evidence is needed to determine exactly why women with diabetes have higher healthcare expenditures compared to men.

The findings of this study are important because they show women have higher incremental costs associated with health services identified as instrumental for diabetes management: prescriptions, office-based visits, and home healthcare services. Prescription medications are needed, not only for managing and controlling diabetes, but also to slow the progression of adverse complications, which can contribute to expenditures when diagnosed. Evidence suggests, for example, that when women have higher healthcare expenses, they often neglect their own care and basic material securities because of medication costs [33, 34]. In a study to identify problems associated with OOP costs among older adults with diabetes, Piette et al. [34] found women more likely to reduce their use of prescribed medications and spend less on basic needs when met with the difficult choice of paying for prescriptions. In addition to prescriptions, office-based visits with providers and procedures are necessary for routine follow-up and clinical assessments (i.e., glycosylated hemoglobin A1c and lipid measurements, along with monitoring of blood pressure, etc.). Furthermore, home healthcare is often required to provide diabetes education (i.e., use of glucometer, examinations of feet, etc.) or to deliver needed equipment and supplies (i.e., glucometers, test strips, lancing devices).

Given the individual and collective importance of various healthcare services in managing diabetes, it is important to determine best practices for lowering associated costs, which will likely lead to improvements in health outcomes among adults with diabetes. In this regard, interventions could be useful directly, to obtain a better understanding of diabetes-related differences in health outcomes by sex and inform policy, clinical practice, and future research, and indirectly, to help minimize sex-specific differences in outcomes, including costs. Interventions that focus on sex differences in diabetes education, self-management, and lifestyle behaviors such as medication adherence, independently or in concert, are critical for helping to improve overall health and outcomes and reducing the need for additional healthcare services that could result in higher costs. These interventions could also stress the importance of preventive services that can be beneficial in preventing or reducing the onset of comorbid conditions and diabetes-related complications, which can lead to higher differential costs. Cost-benefit analyses could be included as a component in the intervention to address concerns related to diabetes care costs and brainstorm approaches for accessing resources to help minimize these expenses. Provider offices, patient homes, and community settings are all possible locations for delivering the interventions and should be determined based on the needs of the adult participants. In addition, as technology continues to advance, interventions that utilize telehealth and telemonitoring are essential in helping to reduce expenditures associated with seeking health services by reducing the need, for example, for excessive outpatient and emergency room visitations.

There are several limitations that must be mentioned. First, the research design for this study is cross-sectional, which limits the ability to infer cause and effect relationships. Second, there are potential confounders that were not controlled for including healthcare utilization and access to care, self-management practices including medication adherence, diabetes knowledge, and social support. Additionally, we did not account for type or duration of diabetes nor the types and quantities of medications prescribed. Third, the findings of this study are based on self-reported data; therefore, the possibility for under- and/or over-reporting and estimating is possible. However, self-reported diabetes has been shown to be reliable with 98% agreement with a medical record [35]. Fourth, these analyses focus on OOP and total direct costs and do not consider indirect costs, which might differ between men and women with diabetes.

The results of this study are important and provide new information about differences in healthcare expenditures between men and women with diabetes. In this nationally representative sample of adults with diabetes, women were found to have significantly higher healthcare expenditures for care compared to men, after adjusting for confounding factors. These findings suggest a need for policies that support the prevention of diabetes and its associated complications and focus on the reduction of OOP and total direct expenditures associated with inadequately controlling and managing diabetes in adults. Also, clinicians must ensure clinical guidelines are followed to reduce the development of comorbid conditions and delay the onset of adverse complications, both of which can influence costs. In addition, providers should proactively assist patients by facilitating their access to resources such as medication assistance programs and patient navigators, who consistently and frequently work directly with patients to determine ways for reducing costs associated with care. Finally, the literature evaluating sex differences in healthcare expenditures among adults with diabetes is limited. Further research is needed to understand the drivers of these differences and to identify potential causes associated with higher costs in women with diabetes.