Background/Purpose: In Europe, 70- 80% of all healthcare expenses are attributable to chronic diseases and a large part of these are musculoskeletal conditions (MSKC).Having >1 disease (multimorbidity) is likely to increase the costs of care but little is known about the association of multimorbidity and health care costs (HCC) and the specific role of MSKC as co-morbid disease in this association. We aimed to explore1) whether the number of morbidities has an important association with costs of care and 2) whether MSKC have an additional impact when occurring as co-morbid disease.

Methods: In a Dutch cross-sectional study, 8904 subjects (>18 years) completed a questionnaire on sociodemographic and lifestyle factors, self-reported physician-diagnosed diseases (MSKC, diabetes, cardiovascular diseases, cancer, migraine, respiratory, skin, mental and bowel conditions) and health care use (general practitioner, rheumatologist, orthopedist, physiotherapist, other specialists, hospitalization in regular/academic hospital or nursing home, home care and domestic help).The total HCC were computed for a 3-months period using reference prices of the Dutch manual for pharmaco-economic healthcare evaluations 2010, accounting for inflation by Consumer Price Index. Missing values were imputed by means of multiple imputation. To deal with skewness, zero-inflated negative binomial regression (ZINB) models were computed to assess 1) the association of number of diseases and HCC and 2) which disease or combination of diseases (in- or excluding MSKC) was associated with the largest increase of HCC using the healthy as reference. Models were adjusted for age, gender, education, origin (western vs. non-western), smoking status and BMI. For each of the different subgroups, based on number or combination of morbidities, raw and predicted 3-months HCC were presented for male/female patients. Predicted HCC were derived from the ZINB-models.

Results: SKC occurred in 1766 cases (20%).Multimorbidity was present in 1722 cases (19%). HCC increased steeply with increasing number of morbidities (e.g. HCC for 1 morbidity were approximately 2 times higher than for the healthy, exp(β)=1.8 [1.7-2.0]). Compared to any other condition, MSKC was associated with higher HCC when occurring alone or when occurring as co-morbid disease. For example, when 2 morbidities other than MSKC were co-occurring HCC were approximately 2 times higher than in the healthy (exp(β)=2.2 [2.0-2.7]) while when one of the two morbidities was MSKC the costs were 3 times higher than in the healthy(exp(β)=3.0 [2.7-3.7]) (Table 1). Conclusion: The total HCC increase with increasing number of morbidities. MSKC are accountable for higher costs of care compared to other diseases independent of the number of morbidities. These important findings deserve the attention of policy makers, especially by prioritizing MSKC in healthcare budgets.