Differences in medical specialist utilization among older people in need of long-term care – results from German health claims data

Higher age is associated with increasing morbidity and higher health care utilization [1]. In industrialized countries, the majority of the elderly aged 60 years and over suffer from at least one chronic disease [2‐4]. Consequently, elderly people show higher health care need and higher utilization of medical care than the overall population [5, 6].

However, previous studies have shown that those elderly who have need for long-term care tend to have unmet health care needs. For instance, elderly who live in nursing homes have less visits to medical specialists such as dentists, orthopedists, ophthalmologists and otorhinolaryngologist [7‐12]. Consequently, if this lower level of medical utilization represents unmet care needs nursing home residents may be at risk of adverse outcomes, i.e. inadequate medication and therapy, avoidable hospitalizations, or falls [13‐18].

The reasons for lower health care utilization of elderly in need of long-term care may lie in their physical and mental limitations. Frailty is an increasing phenomenon in the elderly population [19], and elderly in need of long-term care tend to be more frail than those without need of long-term care [20]. As a consequence, elderly in need of long-term care need more support in activities of daily living [21]. Moreover, elderly people in need of long-term care often suffer from cognitive impairment such as dementia [22‐24]. Both frailty and cognitive impairment can increase the risk of unmet health care needs and perceive barriers to health care because the elderly may have more difficulties to assess their health care need, and to organize health care [25‐28].

In Germany, about 90% of the population (i.e. 73 million people) are members of the statutory health insurance and therefore automatically also members of the statutory long-term care insurance. Only 10% of the population uses private health insurance and private long-term care insurance. Insured persons with physical and mental limitations can obtain long-term care benefits either in cash (to organize informal care by informal caregivers at their home), or in kind (to organize formal care at their home or long-term care in nursing homes). People with higher levels of long-term care obtain higher benefits from the long-term care insurance than people with lower levels of long-term care need. Benefits are granted irregardless of personal or household income or assets. However, benefits are capped, and statutorily insured persons have to pay the remainder if the costs of long-term care services exceed the granted benefits. The goal of these benefits is to provide insured persons with physical and mental limitations with support in activities of daily living. Therefore, long-term care services are distinct from medical services which aim to cure or prevent diseases.

Depending on the degree of physical and mental limitations and the according need for assistance in matters of daily life, care-dependent people are assigned a level of long-term care need. In Germany, until 2016, the level of long-term care need was called Pflegestufe. The Pflegestufe differentiated between the levels 1, 2, 3, and hardship cases. People in level 1 need assistance with daily activities (i.e. personal hygiene, mobility or eating) once a day, whereas people in level 2 need assistance three times a day. People in level 3 need several hours of assistance over the whole day. Hardship cases need at least 7 h of assistance during the day and at least 2 h of assistance during the night. In 2017, the Pflegestufe was replaced by the Pflegegrad which now differentiates between five care grades.

As elderly with a higher level of long-term care need are more limited in matters of daily life including the organization of medical care, they may have a higher risk of inadequate health care. However, although health care utilization of older people has been investigated [5, 29, 30], existing research on health care utilization of older people in need of long-term care is limited [31].

Consequently, the aim of this study is to focus on this particular group and to investigate whether differences in the utilization of 12 medical specialties exist between elderly in need of long-term care and elderly who are not in need of long-term care with a special focus on both the level of long-term care and the care setting. Such differences could indicate perceived access barriers and inequalities in medical care utilization among care-dependent elderly. In the following sections we describe our data source and statistical analysis followed by a description of our descriptive and multivariate findings. The findings are then discussed in the discussion section. Finally a conclusion is drawn.

We gave an overview of medical specialist utilization of elderly people with differing levels of long-term care need. When controlling for differences in morbidity, we found that elderly with higher levels of long-term care were at a much higher risk of having no medical specialist visit than elderly who were not in need of long-term care. This finding applies to 8 out of 12 medical specialties, and the results were similar among the home care and the nursing home setting.

These findings support the importance of differentiating medical care utilization between elderly with and without need of long-term care. Although our descriptive findings showed that older people have on average at least one annual utilization to medical specialists, the picture is more nuanced. Our findings support previous studies that have already shown that elderly in need for long-term care have less visits to medical specialists, especially those living in nursing homes [9, 10]. Our results exceed previous findings showing that increasing levels of long-term care are associated with even lower specialist utilization. Although such kind of formal differentiation between different levels of long-term care need only exists in some countries so far [34‐36], the implications of our findings may be cross-national: People who have limitations in the organization of daily activities may face barriers to medical care. The larger the limitations, the higher the risk of access barriers. One could argue that GPs might substitute the role of medical specialists among elderly in need of long-term care given the relatively high GP utilization of this group [9]. Yet, we controlled for GP utilization but did not find consistent or strong evidence for a relationship between GP visits and specialist visits (see Additional file 4): Having a GP visit either comes with a small decreased probability of not having a specialist visit or is unrelated to medical specialist utilization.

The gradient effects of the level of long-term care may be explained by objective or subjective access barriers. Especially elderly wheelchair users or homebound elderly face objective access barriers because in Germany, it is mostly GPs but not medical specialists who make (nursing) home visits. Subjective barriers to medical care also exist among elderly for instance feelings of shame in case of urinary incontinence [37]. Again other elderly perceive less need because of reduced health expectations. Contrary to younger people, older people perceive ageing as a natural, degenerative process not necessarily requiring a health care visit [38]. Also, they weigh the perceived value of medical care against the physical/mental burden of seeking care [25]. Yet, there is little evidence on the health care seeking process of elderly people in need of long-term care, especially of those living in nursing homes.

Although we found clear gradient effects on the risk of having no medical specialist visit, we did not find suchlike effects on the intensity of received specialist care. On the one hand, the intensity of received specialist care may mostly depend on the severity of disease which is not reflected in our disease categories. On the other hand, claims data do not capture the actual number of visits per person. Multiple visits to the same medical specialist per quarter are documented only as one visit. Only if a patient visits different medical specialists per quarter then each visit is documented. Therefore, the results on the intensity of specialist care have to be interpreted with caution.

Further limitations of our data concern the generalizability and the explanatory power of the models. Despite the fact that the AOK health and long-term care insurance fund covers large parts of the population in Germany, samples from statutory health insurance funds are selective and may not be representative of the overall population [39]. Also, the analyses were based on elderly with a diagnosis in the investigated disease category. Such a diagnosis requires a previous practitioner visit. Consequently, elderly without a practitioner visit, despite a respective disease - were not included in the analyses.

Moreover, the explanatory power of the analyzed models was relatively low ranging from 0.023 to 0.211 (see Additional file 4). Although the McFadden pseudo R² measure generally produces lower coefficients than the R² measure in linear regression analysis [40] our models do not include many of the factors that have been shown to explain health care utilization behavior [41‐44]. For instance, health insurance claims data do not provide socioeconomic or demographic information which might have better explained differences in medical care utilization through differing health beliefs. Given the small range of covariates we did no systematic robustness test of our model.

Our results indicate lower medical care utilization among elderly with higher levels of long-term care. However, although the underlying mechanisms of the findings have not been investigated so far, the findings imply that nursing care personnel and family relatives may not always be able to recognize the need for medical care among older care-dependent people. Furthermore, the findings indicate, that some elderly in need for long-term care may perceive physical or emotional barriers to medical specialists. Depending on the underlying mechanism, these findings may imply that long-term care provider and family relatives should be better trained in the assessment of medical care needs. Furthermore, the organization of and access to medical care should be improved, e.g. by organizing more home visits by medical specialists, and by better cooperation between GPs and medical specialists.