Stratifying healthcare costs using the Diabetes Complication Severity Index

https://doi.org/10.1016/j.jdiacomp.2012.02.004Get rights and content

Abstract

Objective

We aim to determine whether healthcare costs for patients diagnosed with Type 2 Diabetes Mellitus (T2DM) are associated with the severity of diabetes complications as measured by the Diabetes Complication Severity Index (DCSI).

Methods

Retrospective cohort analysis was performed on a 2007 primary care cohort of T2DM patients. The DCSI is a 13-point scale, which comprises 7 categories of complications and their severity levels. Healthcare cost data from 2008 and 2009 were used as primary outcome. Inpatient and outpatient costs incurred for services consumed by patients within the provider network were included. Generalized linear model with log-link and gamma distribution was used to predict healthcare costs.

Results

Of the 59,767 T2DM patients, 2977 (5.0%) deaths occurred and 1336 (2.2%) were lost to follow up. Healthcare cost was strongly associated with increase in DCSI score. Compared to patients without complications, those with more complications (higher DCSI score) had an increased risk of higher healthcare costs. Risk ratio (RR) increased from 1.25 (95%CI: 1.19–1.32) for DCSI = 1 to 1.61 (1.51–1.72) for DCSI = 2; 2.10 (1.91–2.31) for DCSI = 3; 2.52 (2.21–2.87) for DCSI = 4 and 3.62 (3.09–4.25) for DCSI  5. As a continuous score, a one-point increase in the DCSI was associated with a cost increase of 27% (95%CI: 1.25–1.29).

Conclusion

The DCSI score is a useful tool for predicting direct healthcare costs. The DCSI can be used to triage high-risk patients for more focused secondary prevention interventions at primary care level, in a bid to lower overall healthcare costs.

Introduction

In Singapore, the prevalence of diabetes mellitus has risen to a 12-year high of 11.3%, from 8.2% in 2004, and 9% in 1998 (EDCD, 2011). The prevalence of diabetes in Singapore was found to be higher than other Asian countries such as Taiwan (5.7%), Japan (7.3%), and Hong Kong (10.2%) (IDF, 2006). Diabetes mellitus is a complex chronic illness that is associated with multiple long-term complications when control of blood glucose and cardiovascular risk factors is poor. This in turn increases health care costs (American Diabetes Association, 2008, Fu and Qiu, 2009, Menzin and Langley-Hawthorne, 2001, Wang and Geiss, 2009), with the global healthcare expenditure expected to total USD376 billion in 2010, and exceeding USD490 billion by 2030 (Zhang et al., 2010). Diabetes complications have been found to account for more than 35% of direct medical expenditures for this disease (Hogan et al., 2003).

Recognizing the burden of diabetes, the healthcare sector has developed various initiatives to improve compliance with pharmacotherapy, preventive screening, and support for self-care for these patients (Aubert and Herman, 1998, Sadur and Moline, 1999). However, cost-intensive disease management efforts become inefficient when they are directed at the entire diabetes mellitus population, including those at low risk of complications. Since managing healthcare costs and reducing morbidity requires targeting “at risk” patients to the appropriate treatment interventions to ameliorate the comorbidities of the disease (Rosenzweig et al., 2002), it is important to develop tools to systematically stratify patients into the different risk categories.

End-organ complications usually develop simultaneously or consecutively in a patient with diabetes rather than independently (Ramsey et al., 1999). The many combinations of diabetic complications add varying degrees of complexity to the overall care of patients, resulting in different utilization of healthcare services and associated costs. Most studies quantified the impact of individual microvascular, and macrovascular complications on healthcare costs (Joyce and Iacoviello, 2004, Nicholas and Brown, 2002). It is also important to note that complications tend to have a multiplicative effect on healthcare costs but studies utilizing traditional linear regression methods tend to assume an additive effect (Clarke et al., 2003). By using an index that summarizes the burden of several complications into a single-value, we can analyze the concurrent risk of higher healthcare costs associated with diabetes complications. Despite these advantages, few studies have reported the use of an overall diabetes complications severity as a risk stratification tool for healthcare costs (Rosenzweig et al., 2002).

A diabetes severity score, derived from automated large administrative and clinical databases, would give researchers and healthcare administrators an easy and reliable method to stratify patients. It would aid in evaluating the short-term direct costs and consequences associated with diabetes interventions. Scoring patients according to the an overall diabetes severity index would enable easy identification of high-risk patients and triage of patients to specific clinical pathways, and the assessment of the effects that each medical condition has on the overall treatment plan and healthcare costs.

Therefore, we aim to construct a validated clinical tool, the Diabetes Complication Severity Index (DCSI), for patients diagnosed with Type 2 Diabetes Mellitus (T2DM), and to determine whether the severity of diabetes complications is associated with healthcare costs in Singapore.

Section snippets

Study design

This is a cohort study of 59,767 T2DM patients who had received care at least twice in one of the 9 National Healthcare Group (NHG) primary care clinics in 2007. We have defined the baseline visit as the last visit made in 2007. All healthcare costs incurred for services provided by the NHG during the two-year follow-up period from Jan 1, 2008 to Dec 31, 2009, were included.

Besides primary care services, NHG also provides inpatient, specialist outpatient, and 24-h emergency through a network of

Results

The study population comprised 59,767 T2DM patients who had received diabetes care at least twice in the same primary care clinic in 2007. Of these, 1336 (2.2%) patients who were lost to follow-up and 2977 (5.0%) patients who died during the study period were excluded. In total, 55,454 (92.8%) patients were included in the analysis.

The clinical and demographic characteristics of the patients are summarized in Table 1. Of the eligible patients, 31,794 (57.3%) had no complications, 11,486 (20.7%)

Discussion

Diabetes mellitus is a chronic condition with many potential macro- and microvascular complications. The amount of healthcare resources and the corresponding cost needed to care for a T2DM patient with no complications compared with another with complications are vastly different. With a severity index such as the DCSI, patients with varying degree of complications severity could be better stratified so as to customise care interventions based on their expected risk. This study analyzed the

Conclusion

This study shows that the DCSI as a diabetes complication severity score of T2DM patients is strongly associated with healthcare costs, even after adjusting for the other medical conditions and factors. It can be used to identify T2DM patients who need more intensive management to prevent or delay the onset of new complications.

References (32)

  • P. Clarke et al.

    The impact of diabetes-related complications on healthcare costs: results from the United Kingdom Prospective Diabetes Study (UKPDS Study No. 65)

    Diabetic Medicine

    (2003)
  • J. Cohen et al.
  • International Diabetes Federation

    Diabetes atlas

    (2006)
  • Epidemiology and Disease Control Department

    National Health Survey 2011

    (2011)
  • A.Z. Fu et al.

    Health care and productivity costs associated with diabetic patients with macrovascular comorbid conditions

    Diabetes Care

    (2009)
  • P. Hogan et al.

    Economic costs of diabetes in the US in 2002

    Diabetes Care

    (2003)
  • Cited by (21)

    • Incidence and Prevalence of Microvascular and Macrovascular Diseases and All-cause Mortality in Type 2 Diabetes Mellitus: A 10-year Study in a US Commercially Insured and Medicare Advantage Population

      2019, Clinical Therapeutics
      Citation Excerpt :

      The use of diagnosis claims to ascertain microvascular and macrovascular disease in outcomes research studies is well established. One prior study has shown that both the number and severity of diabetes complications (constituents of the Diabetes Complications Severity Index as determined by the presence of ICD diagnosis codes)18,19 were independently associated with increased risk for mortality and hospitalization in a population-based sample of primary care patients with diabetes,20 and have been relied on to estimate costs, and in other types of analyses.21,22 A composite outcome (≥1 claim for any microvascular or macrovascular disease between January 1, 2006, to earlier of December 31, 2015, or death) was also created.

    • The Risk of TB in Patients With Type 2 Diabetes Initiating Metformin vs Sulfonylurea Treatment

      2018, Chest
      Citation Excerpt :

      However, we used the aDCSI score at baseline as a surrogate marker for hemoglobin A1c and adjusted this variable in multivariate analyses to minimize bias from different DM severities. To verify aDCSI as a practical alternative to hemoglobin A1c, further research is warranted to confirm the relationship between the diabetic complication score and hemoglobin A1c level.38 Second, the reasons for initiating metformin and sulfonylurea treatment may have been dissimilar, although they are the two most frequently used initial drugs for T2DM in Taiwan and are contraindicated in advanced CKD.12,14,17

    • Stratification strategy for evaluating the influence of diabetes complication severity index on the risk of hospitalization: a record linkage data in Western Australia

      2017, Journal of Diabetes and its Complications
      Citation Excerpt :

      Costs of health care for people with diabetes with complications have been found to be substantially higher than in those without complications.11,13,14 Recent studies have examined the effect of diabetic complications on health care utilization using the number of complications or DCSI as a continuous variable or a categorical (ordinal) factor in linear regression.13,14,32,33 Although the linear approach is flexible enough to examine the overall pattern of the relationship between the number of complications or DCSI and health care utilization, it may not reflect the underlying probability of the relationship given the conditional nature of subsequent events on prior complication.20

    • Changes in prevalence of diabetic complications and associated healthcare costs during a 10-year follow-up period among a nationwide diabetic cohort

      2015, Journal of Diabetes and its Complications
      Citation Excerpt :

      Noteworthy, patients with aDCSI score of 5 or more only accounted for 3.85% of the whole diabetic cohort but consumed highest healthcare costs than other patients. Similar findings were also found in the Wu et al. (2012) study that patients with aDCSI score of 5 or more only accounted for 1.6% of the whole diabetic cohort but consumed 5.9% healthcare costs. These results emphasize the care needs for this specific subgroup of patients with diabetes.

    View all citing articles on Scopus

    Conflict of interest and ethics statements: All authors are employees of NHG. This manuscript was prepared without any contract or funding from a sponsor. The authors state that the study has been approved by NHG Domain Specific Review Board (DSRB).

    View full text