Statement of principal findings
Almost half of the study population had multimorbidity, with infections (mainly acute upper respiratory infection) the most common acute disease in both sexes and all age groups. The most frequent multimorbidity pattern of chronic diseases was the combination of hypertension and dyslipidemia in adults over 45 years of age.
We observed a decrease in the number of acute diseases recorded as age increased. Nonetheless, in adjusted models female older than 65 who had acute diseases were more likely to have multimorbidity.
Finally, the use of health services was positively associated with a diagnosis of multimorbidity. Living in a rural area decreased the probability of multimorbidity.
Strengths and weaknesses of the study
A major strength of this study is the analysis of a large, high-quality database of primary-care records, representative of a large population. In the context of a national health system with universal coverage, EHR data have been shown to yield more reliable and representative conclusions than those derived from survey-based studies . Another important strength was the inclusion of all chronic and acute diagnoses registered in EHR, which contributed to a more accurate analysis of the association between acute and chronic diseases and of the disease combinations present in multimorbidity in this population. To synthesize the results, we present here only the most frequent combinations. Finally, few studies have incorporated acute diseases in the study of multimorbidity patterns  and none analyzed the relationship between multimorbidity and acute morbidity.
Some possible biases could have influenced our results. First, diseases could be underreported, especially for male of normal workforce age who tend to see their doctors less often than other strata of patients. This effect would diminish in the two oldest age groups because of the retirement age (65 years) in Spain. In patients with multimorbidity, the true incidence of acute disease could be underreported because the GP would place higher priority on the chronic problems in these patients. On the other hand, there could be an over-representation of chronic diagnoses (e.g., hypertension, diabetes, hyperlipidemia, etc.) that are included in the goals/incentives contracts of Catalan PHCT.
The diseases that form part of the CatSalut treatment objectives may be more carefully recorded than other conditions. However, these same diseases are the most prevalent (high blood pressure, diabetes, hypercholesterolemia, smoking, dyslipidemia, ischemic heart disease, atrial fibrillation) and therefore have the greatest impact on population health. The quality-recorders database (SIDIAP-Q) used in this study minimizes the under-reporting of diseases not included in the CatSalut objectives.
Furthermore, the stratified analysis allows more accurate estimation within each age-sex strata and universal access to free health care and medications makes it more likely that patients seeking care will acquire a diagnosis, either acute or chronic [17, 18]. Second, there is no universally accepted criterion for consensus classification of acute and chronic disease. This lack of accurate case definitions impedes the establishment of the true incidence/prevalence of a disease . Finally, a residual confounding cannot be completely excluded, and could occur because of epidemiological factors not considered in this study, such as patients’ socioeconomic status .
Strengths and weaknesses in relation to other studies
The estimated multimorbidity prevalence in our sample is higher than in other European studies [21–24], perhaps because of the analysis of a greater number of diseases in our study than in most other published studies . Nonetheless, the patterns of multimorbidity observed were similar to those observed in other studies .
As in other studies, multimorbidity was more prevalent among female [21, 22, 24]. This could be due to the longer female life expectancy and worse health status, compared to male, differences that are due to both biological and social factors . In addition, sex is a social determinant that influences health status, health behaviours and the use of health services [27–30]. Recent studies, however, suggest a dismantling of this paradigm based on sex-stratified analysis of consultations for common symptoms .
Acute problems are time-consuming for health professional , and therefore should be considered part of the primary care workload. Although current health policy, health care services, and research are all heavily focused on chronic diseases, we must not forget that 41% of primary care visits are motivated by an acute disease . The incidence of acute diseases observed in our study concurs with other reports of acute upper respiratory infection and other health problems of infectious aetiology (acute tonsillitis, cystitis) as the primary reasons for seeking primary care, along with non-infectious diseases such as dorsalgia [33, 34].
Our study observed a higher prevalence of multimorbidity in rural settings. Other studies conducted in rural areas have reported only a greater prevalence of multimorbidity in elderly people [6, 7]. Nonetheless, our adjusted analysis showed that living in a rural area is negatively associated with multimorbidity. This phenomenon could be due to the environmental and sociocultural context and access to both public and private services .
Implications for clinicians and policymakers
Our study considered multimorbidity in patients who received primary medical care, considering all visits and diagnoses (acute and chronic diseases). This approach allowed the identification of vulnerable subgroups in our population. A major advantage of our methodology is the use of data obtained directly from standard clinical practice. Knowing the distribution of acute and chronic diseases by life-stage and sex will help the clinician faced with a particular patient to anticipate disease patterns based on the patient’s sex and stage of life, recognizing that these vary with age. This will encourage the implementation of personalized disease prevention and health promotion activities.
At the level of health policy and health care administration, the organization of services should be reviewed to ensure that continuity and coordination of patient care are guaranteed; current evidence suggests the potential for improvement in this regard .
Unanswered questions and future research
The classification of chronic and acute disease remains unresolved, and there is no consensus on the type and number of chronic diseases that define multimorbidity. A personalized measure to determine the severity of diagnosed multimorbidity is also needed.
If longitudinal studies confirm a higher incidence of morbidity in patients with multimorbidity, evidence-based interventions will be needed to prevent the onset of acute disease. Further studies are needed to study possible genetic and pathophysiological explanations that corroborate the observed multimorbidity patterns.
In-depth analysis of other contextual factors related to multimorbidity is also required, along with studies of the relationship between area of residence and multimorbidity and of the differences in health status that may exist between different territories. Finally, there is a need for the implementation and evaluation of health literacy and self-management interventions to improve patient competence in resolving routine acute diseases, which in turn will decrease the care burden in primary care systems.