The integrated care model of our geriatric ambulatory clinic did reduce the frequency of outpatient visits and the number of drugs prescribed and thus the total annual cost of outpatient care. However, medical expenditures associated with hospitalization increased in the first year for the Geri-OPD patients compared with the non-Geri-OPD patients. In the second year, ambulatory clinic visits, emergency department visits, the frequency of hospitalizations and cost decreased significantly compared with the first year of geriatric integrated outpatient clinic care. The Geri-OPD patients still had more outpatient and emergency visits and higher medical expenditures for every hospitalization than the non-Geri-OPD patients; however, the annual medical costs for outpatient care and hospitalization and the length of hospitalization did not significantly differ between these two populations after two years.
In this study, the dramatic decrease in the number of medical services (outpatient visits) and drugs prescribed reduced medical expenditures in the geriatric integrated outpatient clinic. The older patients who received geriatric integrated outpatient clinic care in our study had reduced health care utilization in the first year (Figure 1), and this was reduced even further in the second year, which was better than the results in other reports in recent years [13, 14]. The overall proportion of patients with multiple doctor visits was as high as 39.4%, according to the National Health Insurance Research Database in Taiwan [23]. The top five reported diagnoses of older people treated at ambulatory care visits were diseases involving the circulatory, respiratory, musculoskeletal, and nervous systems and endocrine disorders [24]. Our geriatric integrated care model involved geriatricians, neurologists, psychiatrists and doctors in physical medicine and rehabilitation, social workers, health educators, dietitians and pharmacists. Thus, we could handle the most common problems of older adults who were frequent users of insurance. The CGA is an important part of our integrated care system at the outpatient clinic, especially in the evaluation of geriatric syndrome and patient-centred care. There is currently no standardized algorithm for acute and chronic care for older people [25].
The average number of outpatient clinic visits of older adults (older than 65 years) in Taiwan was reported to be 26.8 ± 22.7 (mean ± standard deviation) visits per year in 2004 [24]; the number in our study was 66.40 ± 51.45 for the Geri-OPD patients and 35.86±48.53 in the control group. The number of outpatient clinic visits in our study was much higher than that in other reports [26-28]. The high health care utilization was related not only to how complicated the patients’ conditions were but also to the near-total coverage of medical fees by the national insurance system. The latter factor affected the patients’ health care-seeking behaviours. The most common comorbidities in patients with multiple physician visits in Taiwan are type 2 diabetes mellitus (3.68%) and hypertension (3.79%), according to the National Health Insurance Database [23]. In our study, the most common comorbidities were diabetes mellitus (19.2%) and cerebrovascular accidents (20.1%). Our patients had more complicated conditions than the average patient in Taiwan. The reduction in health care utilization was even more obvious in the second year of our study, which indicated that the treatment plans required time to take effect for older patients with complicated conditions. The trend was most obvious in the oldest patients in the population. The number of prescriptions actually increased in the first year, which may have been secondary to newly diagnosed problems, such as geriatric syndrome. Geriatric syndrome was frequently missed and was considered normal ageing before the implementation of the CGA. Therefore, increased health care utilization was needed in the first year to treat newly diagnosed medical problems and modify drug use. The use of the CGA in outpatient services has been less studied. One possible reason is that the CGA and subsequent formation of an individualized plan are time consuming. We tried to separate the CGA and the plan formulation process into the first 2 or 3 visits in our clinical practice, and this helped us to identify major and potential problems in an efficient way. According to a previous review article, the performance of the CGA in the clinic had no statistical effect on survival, but one recent randomized controlled trial showed a beneficial effect on frailty after 2 years in patients with very complicated conditions (age ≥ 75 years, ≥ 3 current diagnoses, and ≥ 3 hospitalizations during the one year prior to study inclusion) [29, 30]. Our study provides evidence that performing a CGA in the outpatient clinic for patients with complicated conditions can reduce their high health utilization for 2 years.
Health care utilization was reduced in the subgroup of Geri-OPD patients with a high number of comorbidities in our study (Figure 3). Approximately 34% of the patients in our study had a high number of comorbidities (CCI≥2). Higher numbers of comorbidities are correlated with an increased cost of hospitalizations and high economic burden, which was also found in our study. Librero reported that patients with more comorbidities have a longer length of hospital stay, higher mortality, and higher readmission rates [31]. A higher number of comorbidities is likely to lead to more complications during admission, which results in higher medical costs for each hospital stay. Multidisciplinary interventions can reduce hospital admissions and falls in older adults and increase patient satisfaction with health care services, but institutionalization and mortality rates might not decrease [32-34]. In our study, the subgroup of patients with more comorbidities had higher health care utilization than the subgroup of patients older than 80 years (Figure 4). The high number of comorbidities was associated with high health care utilization in our study. However, the average age of all patients and of the two subgroups was older than 80 years, which may underestimate the effect of ageing on health care utilization.
Initially, the average number of medications used by our patients was more than 20. The number of drugs used by Geri-OPD patients after geriatric integrated outpatient clinic care was significantly less than that of non-Geri-OPD patients. Nevertheless, all of the patients in this study still took more than 15 kinds of medications, even after the second year. It was difficult to reduce the number of drugs used in patients with multiple comorbidities. Thus, we focused on the prevention of potentially inappropriate prescriptions. In several studies and reviews, polypharmacy and inappropriate prescribing had an adverse effect on older people due to the higher risk of falls and drug-related harm [35-37]. From 2001-2004, 19.1% of patients older than 65 years who were covered by Taiwanese National Health Insurance had an inappropriate medication prescription according to the Beers criteria [38]. Older people taking inappropriate medications have significantly more ambulatory care visits, emergency department visits and hospital admissions [38]. A prospective study including 6,666 adults aged older than 50 years in Ireland revealed that polypharmacy (>4 medications) was associated with the number of falls in older adults if antidepressants or benzodiazepines were included [39]. A previous study revealed that 50% of older adults take one or more medications that are unnecessary and that having a clinical pharmacist on the multidisciplinary team could help reduce drug numbers [40]. Polypharmacy could be attributable to the presence of multiple chronic diseases and to patients visiting multiple ambulatory clinics because Taiwan’s health insurance does not restrict health care utilization by any individual. In our geriatric outpatient integrated care system, unnecessary medications were discontinued after the treatment goal was set, and the patient’s functional status was considered. This intervention could slow down the vicious cycle of comorbidities, multiple ambulatory clinic visits, polypharmacy, emergency department visits, and hospitalizations.
An integrated care model for older people was discussed in 1983 by Albert and included acute care units, rehabilitation day hospitals, nursing homes, outpatient clinics, and home care [41]. Compared to traditional outpatient clinics, integrated outpatient clinics can decrease acute care utilization, reduce medical costs and decrease subspecialty clinic use, as was the case for the Collaborative Assessment and Rehabilitation for Elders (CARE) Program in the United States, which was designed for chronically ill older adults who did not meet the indications for inpatient rehabilitation [26-28, 42]. Geriatric evaluation can reduce functional decline without increasing medical cost [42-45]. It is difficult to efficiently develop appropriate guidelines for caring for older people with several comorbidities [46, 47]. In the preliminary data of one study that embedded geriatric service into primary care by providing on-site consultations with a geriatrician and geriatric nurse case manager, the mean number of subspecialty clinic visits (7.4 ± 9.8) declined significantly during the first year after enrolment and after the second year [26]. Fragmented care was evident in that study. In our efficient geriatric integrated care model, which was provided in the outpatient department of the hospital, we reduced the health care utilization of the oldest group of patients, who had complicated conditions and needed a very high number of clinic visits per year.
Integrated health care for older adults is currently an important issue and will continue to be important in the future. Families and well-trained caregivers are an important part of a comprehensive integrated care system. Geriatric clinics that include a range of health services can provide older patients with more convenient medical services [48]. The health care expenditures of healthier older people are similar to those of less healthy people despite their longer life expectancy [49]. Our care model reduced unnecessary medical services and saved medical resources. Polypharmacy and the high frequency of outpatient clinic visits remain serious problems that consume medical resources in our society. A change in health care policy and intervention can improve polypharmacy problems, as evidenced by Japan’s experience and in other studies [50, 51].
This retrospective study has some limitations. There may be patient selection bias in that we did not stratify the patients according to the number of comorbidities. The variation in different comorbidities could affect the amount of medical utilization. The included patients came from only one hospital. However, we selected patients from other internal medicine departments who were similar in age and comorbidities to be the control group and were matched by age and sex. This statistical method used in this study reduced the selection bias. Future studies of the geriatric integrated outpatient clinic approach may focus on its role in caring for patients with multiple comorbidities and for the oldest people and in providing late life care.