Healthcare utilization is substantial for patients with type 2 diabetes in primary care: A patient-level study in a Swedish municipality

Introduction and methodology

The increasing incidence and prevalence of diabetes mellitus worldwide conveys a substantial economic impact on society [1]. In 2001, a cross-sectional observational study was performed at the Primary Health Care (PHC) Centre in the municipality of Ödeshög, Sweden, aimed at exploring the healthcare utilization attributable to patients with type 2 diabetes compared to patients without diabetes at the PHC level in a defined population. Patient-level clinical costing was used for calculating the healthcare cost for each individual patient. Data on morbidity and utilization of healthcare resources were obtained from the computerized patient record and economy data from the general ledger. Eligible for the present study were patients diagnosed with diabetes mellitus, using the code number for diabetes mellitus according to the ICD-10 classification, at a visit to a GP at the PHC centre and/or patients that had made one or more visits to a nurse specially trained for diabetes care, during the study period.

Results

During 2001, 4025 subjects (72%) in the municipality made at least one healthcare visit at the PHC centre. The cost for a GP consultation was on average about EUR 100. The cost for a nurse visit ranged from EUR 20–60, depending on the type of visit.

The average annual cost per patient was EUR 236 (SD 306). The diagnosis that was found to be associated with the highest utilization of PHC resources was diabetes mellitus. The annual mean cost for healthcare consumed by patients with this diagnosis was EUR 841 (SD 2230). The diabetic patient group was about 7% of the total patient stock and utilized about 18% of the total resources.

The distribution of the healthcare resources utilized within the entire study population, subjects ≥40 y without diabetes mellitus, and subjects ≥40 y with diabetes mellitus, is shown in Table I. It emerges that the average healthcare consumption among patients with diabetes was higher than in patients without diabetes in most resource items provided by the PHC centre. The difference in healthcare utilization was most substantial in the resource labelled “home care by nurse”, where consumption was about 13 times higher among diabetic patients compared to non-diabetic patients ≥40 y (p<0.001).

Table I.  Characteristics of healthcare utilization at the Primary Health Care Centre of Ödeshög, 2001.

		All patients n=4025		Patients without diabetes ≥40 y n=2060		Patients with diabetes ≥40 y n=277
Resource item	Unit price (€)	n	Cost per patient	n	Cost per patient	n	Cost per patient	p
Visits to GP (not planned)	89	4653	103	2294	99	315	101	0.221
Visits to GP (planned)	106	1960	51	1303	67	306	117	<0.001
Home visit by physician	112	226	6	164	9	61	25	0.076
Home care by nurse	32	5556	45	2005	31	3443	402	<0.001
Visits to district nurse	23	4563	26	2413	27	634	52	0.023
Visits to nurse	22	2047	11	1155	12	222	18	0.028
Visits to nurse for diabetes care	45	547	6	0	0	543	88	<0.001
Visits to physiotherapist	28	2723	19	1652	22	119	12	0.366
Visits to occupational therapist	55	394	5	282	8	58	12	0.403
Home visit by occupational therapist	64	329	5	244	8	79	18	0.018

Analyses were made with linear regression adjusted for age and gender.

The six most costly patients were all diagnosed with insulin-dependent diabetes mellitus and were also suffering from several other complicating conditions such as heart failure or psychiatric disorders. The reason why these patients were so “resource demanding” was that they were all unable to administrate their insulin injections themselves due to co-morbidity. Thus, these patients were in need of daily home care–-sometimes two to three times daily.

Discussion

Although it is well known that type 2 diabetes is often associated with high co-morbidity [2], the subsequent higher healthcare consumption in PHC has, to our knowledge, not been explored before. However, in our study, we have been able to account for resources consumed by patients with diabetes in PHC in a more complete way by accounting for a variety of healthcare resources provided in PHC such as district nurses and home care. It emerges from Table I that the most important difference in healthcare utilization between the average PHC patient and patients with diabetes was the consumption of home care, which was about 13 times higher in the latter group.

In the CODE-2 study [1], it was estimated that the mean annual cost per patient for all kinds of ambulatory care was EUR 813. In our study, confined only to PHC, the corresponding figure was EUR 841. We have not been able to account for ambulatory healthcare use at other PHC centres or hospital outpatient settings. Our interpretation is that the data from the CODE-2 study of expenditures for ambulatory care may be an underestimation of the true costs of type 2 diabetes.

In conclusion, our study of a defined population that included the vast majority of patients with type 2 diabetes shows that the total healthcare resources utilized in PHC by patients with diabetes mellitus were substantial. Furthermore, the utilization of home care in subjects with diabetes was considerable and will, if not accounted for, lead to an underestimation of the true economic impact of diabetes at the PHC level.