Multimorbidity of chronic diseases and health care utilization in general practice

The presence of multiple coexisting chronic diseases in individuals and the expected rise in chronic diseases over the coming years are increasingly being recognized as major public health and health care challenges of modern societies [1‐6]. Individuals with multiple conditions are presumed to have greater health needs, more risk of complications, and more difficulty to manage treatment regimens. At present the main health care model is disease-focused rather than person-focused. Therefore involvement of several different health care providers in managing multiple disorders is inevitable and often results in competing treatments, sub-optimal coordination and communication between care providers, or unnecessary replication of diagnostic tests or treatments [3, 7, 8]. Hence, the common belief is that persons with multiple diseases have high rates of health care utilization and this is confirmed by some international studies [4, 9‐14]. However, till now there is only limited information on health care utilization patterns related to multiple disorders in the Netherlands.

Primary care based registers represent a valuable source to study the relationship between multimorbidity and health care utilization. The general practitioner is usually the first health care professional confronted with patients’ health problems. Studies exploring care utilization in primary care registers showed that individuals with multiple chronic conditions had more contacts with general practice than those with single conditions [4, 11‐13]. An important question is whether such an increase in contacts is equal for each additional disease or whether the increase levels off from a certain number of diseases or for specific combinations of diseases. During a general practitioner visit several overlapping health conditions may be discussed. It is also possible that the increase in the number of contacts increases with each extra disease, due to competing treatment demands or polypharmacy [15]. More insight in the extent of the increase in contacts informs about the burden of multimorbid patients on health resources and may assist in planning and improving (the organization of) health care services.

The objective of this study was to examine the relationship between having multiple diseases and the number of contacts with general practice. We were specifically interested in the development of the number of contacts per additional disease and for specific combinations of chronic diseases.

Over a fourth (26%) of the patients had multiple chronic diseases (15% had two diseases, 7% had three diseases, 3% had four diseases, and 1% had five or more diseases), 30% had one chronic disease, and 44% had no chronic disease.

The present study shows that patients with multiple chronic diseases had more contacts with general practice, more medication prescriptions, and more referrals to specialized care than patients with one or without any chronic disease. The number of contacts increased linearly with the number of chronic diseases for all types of contacts in general practice. However, the number of contacts per disease decreased with the number of diseases. In line with this finding, almost all patients with comorbid diseases had a lower observed number of contacts than would be expected on the basis of contact frequencies for each of the diseases separately.

Complex morbidity requires more diverse and intensive care [25], which likely explains the higher contact frequency among patients with multimorbidity. The finding of a lower number of contacts per disease is not so easy to interpret. A first explanation may be efficiency in treatment by the general practitioner; related health problems may be managed concurrently. Treatment or self-management strategies for diseases may overlap to a certain extent, and treatments may affect multiple diseases favourably [26]. Secondly, physicians and patients may also prioritize health problems, for instance to retain or reach an adequate level of patient’ well being or functioning [27, 28]. Consequently, treatment for patients with multiple diseases may be suboptimal and chronic diseases may receive less attention than needed. Management of chronic diseases usually takes place in accordance with disease-specific guidelines, which pay only minor attention to treatment of patients with comorbidity, especially for diseases that are not related [29, 30]. When comorbidity of diseases represents part of the same overall pathophysiologic risk profile or has overlapping treatment and self-management strategies (concordant diseases) [15, 26], such as diabetes and coronary heart disease, a lower number of contacts might be expected. Management and treatment of concordant diseases generally affect the status of both diseases favourably [26]. However, for disease pairs that are not associated (discordant diseases) such as stroke and osteoarthritis the observed number of contacts was also lower than expected [15, 26]. This corresponds with earlier findings that non diabetes-related comorbidity increases the health care demand as much as diabetes-related comorbidity [31]. A final explanation for a lower number of contacts per disease is that multimorbidity has a great impact on the balance of use of services between primary care and specialist physicians [9]. Two recent reviews concluded that having multiple diseases leads to a rise in specialized care, such as the utilization of specialist physician services, hospital admissions, and the number and length of hospital stays [1, 14]. For a more comprehensive understanding of healthcare utilization for multimorbid patients, we should also look at the contact rates in specialised care. Specialists dominate the care of people with high burdens of morbidity because of the multiplicity of disease types. Therefore, substitution from primary care to specialised care may have occurred.

Obviously, the rise in the use of health care resources for patients with multiple diseases has consequences for the current and future burden of patients with multimorbidity in general practice, but the finding that per disease the number of contacts is lower may imply a more optimistic tendency. However, it should be noted that it is important to get insight into quality of care. Lower contact rates per disease may also indicate undertreatment. Quality and coordination of care for patients with multiple diseases is a concern since most treatments and guidelines are disease-specific [25, 29, 30]. Currently case-management programmes are being developed and evaluated worldwide for patients with multiple diseases, but it is still largely unknown what constitutes optimal care for multimorbid patients [10, 32]. When quality of care is found to be low among patients with multiple diseases, a decreased number of contacts per disease is an undeserved trend. Underlying reasons for lower contact rates per disease must be explored in future studies.

Our findings are in line with other European studies exploring the relationship between the number of diseases and healthcare utilization in primary care [4, 11‐13]. German and English studies show that primary care utilization more than doubled for patients with multimorbidity (Germany mean 36.3 contacts per year, England 9.4 consultations per year) compared to those who are not multimorbid (Germany 15.9 contacts per year, England 3.8 consultations per year) [12, 13]. As shown by the large differences in contact rates between these studies and with contact rates in our study, not just the number of diseases determines the number of contacts in primary care. The definition of contact rates and probably also accessibility of health resources affect the mean contact rates. For example, prescribing medication is included in our definition of a contact but not for the study in England [12]. This limits direct comparison of contact rates between countries. The reason that prescribing medications was included in our study is that it is actually a combination of prescribing medication and a telephone consultation, because questions about the medication or side-effects are very often discussed. Prescribing medication with or without telephone advice on medication issues cannot be distinguished in this general practice registration. Therefore, including the category of prescribing medication leads to an overestimation whereas excluding this category leads to an underestimation of general practitioner contacts with patients.

The German and English studies also confirm the linear increase of the number of chronic diseases with total contacts, a higher number of medication prescriptions, and a higher number of referrals for patients with multiple diseases [4, 12, 13]. The number of contacts per disease was not studied before. Our study is of interest since our findings showed lower contact rates per disease with an increased number of diseases and lower contact rates than what would be expected for specific combinations of diseases. For disease combinations costs in primary care were studied in another English study [33]. This study showed that the costs of treatment for most combinations of diseases did not differ from costs of two patients each with only one of the diseases. In total 12% of the combinations was cost-limiting, this was mainly observed among people over 60 years. Compared to our findings where almost all combinations of diseases showed lower contact rates than expected, cost-limiting conditions in the Brilleman study were less frequent. Moreover, about 7% of the combinations were cost-increasing and this was especially true for depression in combination with physical comorbidities (diabetes). It is not exactly clear why this differs from our results based on contact rates.

Main strengths of this study are the availability of data on diagnosed chronic diseases and the use of a large nationally representative sample of general practices. However, by using disease counts to define multimorbidity all chronic diseases contribute equally, independent of their severity or prognosis [34]. We noticed a large variation in contact rates between patients (shown by large standard deviations in Table 1), which may be explained by differences in the severity of health problems. It is likely that most severe health problems lead to the highest contact rates. For future research, it is interesting to study the variation in contact rates and identify the characteristics of patients with the highest care utilization. Moreover, disease counts do not account for differences in relationships between diseases. Concordant and discordant diseases are equally summed, while the impact on the patient may differ and be lesser or greater than the simple sum [35]. Although the use of disease counts has some limitations they are most widely used in multimorbidity research and they perform equally well compared to two other multimorbidity indices in determining the relationship with health care utilization [34]. Internationally there is a lot of variation in the number and type of chronic diseases that are considered in multimorbidity research and our selection also differed from others [6, 36]. Generally speaking, the more chronic conditions are included the more patients with multimorbidity will be found. We presume that an important part of chronic morbidity is included in our selection of diseases. Furthermore, by using registration data from general practices the number of contacts for patients treated by specialists is not taken into account. The dataset only contained information about referrals to specialized care and no records of the number of contacts with specialists. As patients treated by specialist are mostly complex cases, we may assume that they usually have higher health care utilization.

Since the majority of older people have multiple diseases and their number is rising, it is important to get more insight in their health care utilization patterns. This study shows that the number of contacts in general practice increased linearly with the number of chronic diseases, thus multimorbid patients account for a high proportion of the healthcare workload. With the expected rise in multimorbidity in the coming decades, this requires more extensive health resources. Furthermore, the explanations for the decrease in contacts per disease should be explored. In case of undertreatment or low quality of care for patients with multiple diseases, advances may even lead to a further increased use of health resources in the future. In conclusion, health systems should be prepared for the future increase in utilisation of health services.