Exploring the completeness of comprehensive geriatric assessment in relation to care outcomes in hospitalized older patients

Abstract

As a population ages, performing comprehensive geriatric assessment (CGA) is necessary to help clinicians manage older adults' conditions and to prevent or delay complications. CGA is now known as the best method by which to improve care outcomes and quality of life in older adults, and it requires a multidisciplinary approach. The aim of CGA is to identify older adults' needs, develop individual care plans, and improve care outcomes.1 Care outcomes are the results of treatment. They're also a key performance indicator for hospitals. The most common negative key performance indicators related to older adults are a long hospital stay, dying in the hospital, and unplanned readmission.2 The implementation of CGA may help control the care outcomes of older adults, as one of its main functions is to improve these outcomes. However, previous studies have shown inconsistent results related to CGA—although some have indicated that CGA improves care outcomes, others have found that CGA can worsen them. Specifically, many studies have reported that the implementation of the CGA model can reduce lengths of stay, readmission rates, and in-hospital mortality among hospitalized older adults, especially in the geriatric ward.3-6 However, others have reported these three care outcome parameters increased following the implementation of the CGA model.4-7 The reason for these inconsistencies requires further investigation. For example, it's possible that the method of CGA implementation varied among studies. As such, investigation is needed into how CGA is being conducted, especially the completeness of the CGA tool used; this is integral to achieving a comprehensive diagnosis and identifying appropriate interventions. Accordingly, this study was conducted to evaluate the completeness of the CGA tool and its relationship to care outcomes among older adults—especially readmission rates, lengths of hospital stay, and in-hospital mortality. Methods Investigators retrospectively evaluated older adults' medical records in one of the first—and largest—referral hospitals in Indonesia. To evaluate readmission rates, patients' medical records were reviewed for 30 days after discharge from the hospital. Included in this study was a sample of 222 hospitalized adults age 60 years or older who had a minimum of two diseases according to medical diagnosis. Older adults who were admitted or transferred to the ICU and those who were hospitalized for less than 24 hours were excluded from the study. The selection technique applied was stratified random sampling. Multiple logistic regression analyses were used to look for the predictor influence of the dependent variables of 30-day readmission and in-hospital death and examine the predictors of length of stay. Geriatric medical teams for participants performed various assessments to evaluate the completeness of the CGA. The researchers created this completeness evaluation, which comprised 4 domains (physical health, functional status, psychological health, and socioenvironmental status) broken into 21 items. The total completeness score ranged from 0 to 21. The data were then converted into percentages to facilitate analysis and evaluation. This evaluation utilized the observational checklist instrument, which had been tested for validity and reliability. Researchers conducted a pilot study to test this instrument and found it to be valid and reliable. It received a score of 0.89 using the content validity index (CVI). This CVI was reviewed by eight experts from Indonesia and Taiwan, including a geriatrician, a geriatric nurse, and gerontological nursing lecturers. The reliability test used an internal consistency and interrater reliability test. The internal consistency test analyzed by Kuder-Richardson Formula 20 with a result of 0.73 means the instrument is reasonably reliable in producing a consistent score. Results Completeness of CGA The overall average completeness of the CGA performed on the included patients was 68.19% (SD, 7.85). The CGA was more likely to be fully complete when conducted on the geriatric ward (71.99%; SD, 8.54) than on the nongeriatric ward (64.52%; SD, 4.85), with a significant difference (t = −7.99; P = .000). Table 1 presents a description of the CGA's completeness in terms of the total score and the score for each of the four evaluated domains. Table 1: - Description of the completeness of CGA (N = 222) Variables Mean ± SD Median (Range) 95% CI Completeness of CGA 68.19 ± 7.85 66.67 (52.38-85.71) 67.15-69.23 Physical health 84.73 ± 5.39 88.89 (77.78-88.89) 84.02-85.44 Functional status 39.34 ± 12.84 33.33 (33.33-66.67) 37.64-41.04 Psychological health 12.01 ± 25.68 00.00 (00.00-66.67) 8.62-15.41 Socioenvironmental status 86.04 ± 10.16 83.33 (50.00-100) 84.69-87.38 Looking at each domain in greater detail, only socioenvironmental status was completely evaluated by the medical care team. In each of the three other domains, at least one item wasn't examined at all. A detailed description of each item evaluated within the four domains is given below, and the data are presented in Table 2. Table 2: - Data distribution of the completeness of CGA for each item by ward type (N = 222) Type of ward Domain (mean ± SD)/Item, n (%) Nongeriatric (n = 113) Geriatric (n = 109) Total X2/U P Physical health 83.19 ± 5.58 86.34 ± 4.69 84.73 ± 5.39 4410 .000∗ Chief complaint 113 (100) 109 (100) 222 (100) - - Medical history 113 (100) 109 (100) 222 (100) - - Healthy behavior history 55 (48.7) 84 (77.1) 139 (62.6) 19.1 .000∗ Physical assessment 113 (100) 109 (100) 222 (100) - - Multimorbidity 113 (100) 109 (100) 222 (100) - - Polypharmacy 113 (100) 109 (100) 222 (100) - - Nutrition status 113 (100) 109 (100) 222 (100) - - Balance 0 0 0 - - Risk of fall 113 (100) 109 (100) 222 (100) - - Functional status 34.22 ± 5.38 44.65 ± 15.86 39.34 ± 12.84 4231 .000∗ Activity of daily living 113 (100) 109 (100) 222 (100) - - Instrumental of activity daily living 3 (2.7) 37 (33.9) 40 (18) 36.77 .000∗ Mobility 0 0 0 - - Psychological health 1.77 ± 10.77 22.63 ± 31.71 12.01 ± 25.68 4231 .000Re Cognition 3 (2.7) 37 (33.9) 40 (18) 36.77 .000∗ Delirium 0 0 0 - - Mood 3 (2.7) 37 (33.9) 40 (18) 36.77 .000∗ Socioenvironmental status 83.33 ± 9.71 88.84 ± 9.9 86.04 ± 10.16 4512.5 .000∗ Marital status 113 (100) 109 (100) 222 (100) - - Guardian 113 (100) 109 (100) 222 (100) - - Caregiver 106 (93.8) 108 (99.1) 214 (96.4) 4.45 .066 Environmental support 108 (95.6) 108 (99.1) 216 (97.3) 2.59 .213 Financial situation 109 (96.5) 104 (95) 213 (95.9) 0.16 .745 Social activity 16 (14.2) 43 (39.4) 59 (26.6) 18.18 .000∗ ∗Correlation is significant at the alpha level (.05). Care outcomes As shown in Table 3, compared with the older adults on the nongeriatric ward, those on the geriatric ward experienced a lower rate of readmission (4.6% versus 8%) but a higher rate of in-hospital death (13.8% versus 2.7%). In addition, the older adults hospitalized on the geriatric ward stayed in the hospital significantly longer than did the older adults on the nongeriatric ward (U = 4854; P = .006). Table 3: - Care outcomes in relation to ward type Type of ward Care outcomes Total Geriatric Nongeriatric X2/U Score P 30-day readmission, n (%) 0.58 .448 Yes 14 (6.3) 5 (4.6) 9 (8) No 208 (93.7) 104 (95.4) 104 (92) In-hospital death, n (%) 7.76 .005∗ Yes 18 (8.1) 15 (13.8) 3 (2.7) No 204 (91.9) 94 (91.9) 110 (97.3) Length of stay, median (range) 8 (2-62) 10 (2-34) 7 (2-62) 4854 .006∗ ∗Correlation is significant at the alpha level (.05). Completeness of CGA in relation to care outcomes The length of hospital stay care outcome showed different results from the other two outcomes as shown in Table 4. In general, there was a significant relationship between the completeness of the CGA and the length of stay (r = 0.15; P = .028). When evaluating by ward type, a significant relationship between the two parameters was observed for patients on the geriatric ward (r = 0.19; P = .041) but not for those on the nongeriatric ward (r = −0.081; P = .393). In other words, the length of hospital stay was significantly correlated with the completeness of the CGA for the total sample and for the geriatric ward. Both correlations were positive and moderate, meaning that the more complete the CGA was, the longer the hospital stay. Table 4: - Completeness of CGA in relation to care outcomes Completeness of CGA Care Outcomes Total (n = 222) Geriatric (n = 109) Nongeriatric (n = 113) Mean ± SD t/r (P) Mean ± SD t/r (P) Mean ± SD t/r (P) 30-day readmission 1.13 (.272) 0.92 (.361) −0.70 (.483) Yes 66.67 ± 4.94 68.57 ± 7.22 65.61 ± 3.18 No 68.29 ± 8.00 72.16 ± 8.59 64.42 ± 4.97 In-hospital death −0.33 (.739) 0.89 (.372) 0.95 (.346) Yes 68.78 ± 8.68 70.16 ± 8.72 61.90 ± 4.77 No 68.14 ± 7.79 72.29 ± 8.52 64.59 ± 4.85 Length of stay 0.15 (.028)∗ 0.19 (.041)∗ −0.08 (.393) ∗Correlation is significant at the alpha level (.05). Correlation between completeness of CGA and care outcomes, controlling for other variables The multiple logistic regression analysis found no significant results related to readmission. Only the type of ward had a significant influence on in-hospital death (odds ratio [OR] = 7.28; P = .004). This indicates that older adults hospitalized on the geriatric ward had a probability of death that was seven times higher than those hospitalized on the nongeriatric ward. Further, analysis indicated that the only predictor of length of stay was comorbidity. However, upon examining this by ward type, the completeness of the CGA had a significant impact on and predicted length of stay only for patients on the geriatric ward (B = 0.02; t = 2.08; P < .05). This result indicates that every change by one unit in the completeness of the CGA increased the length of stay by 2 days. Study authors hypothesize that when a CGA is more complete, more problems are identified, necessitating further medical treatment; accordingly, providing a comprehensive intervention to manage all the identified problems might result in a longer length of stay.6,8 Discussion Completeness of CGA The completeness of the CGA was defined as how comprehensively the geriatric medical team assessed participants during hospitalization. The analyses conducted in this study indicated that, on average, the reviewed CGAs were only slightly more than 50% complete. Unfortunately, this result can't be compared with other studies because no previous studies have evaluated CGA completeness. Moreover, this result doesn't correspond to the concept of CGA, which is the primary recommended intervention for geriatric wards.4 Although policy requirements are in place regarding the use of CGA as part of the provision of geriatric services, the percentage of CGA completeness on the geriatric ward alone was only about 72%. A comprehensive CGA is vital to establish diagnoses and ensure that the interventions provided are appropriate and in accordance with each patient's needs. The results also showed that CGA wasn't optimally implemented in the hospital setting, as some items were frequently skipped. As such, other medical teams—especially nurses—must contribute to the CGA, as they spend a significant amount of time with patients; their input could reduce missing data, making the CGA more comprehensive and complete, enhancing care outcomes for older adults. Care outcomes 30-day readmission. Based on a review of all the older patients who were readmitted within 30 days after discharge from the hospital, individuals on the nongeriatric ward had twice as many readmissions as did those on the geriatric ward. Previous related studies found that patients who had been treated in the CGA group had fewer readmissions.3,5 In terms of the relationship between the completeness of the CGA and readmission, the total sample score of the average number of older adults who were readmitted was lower than that of patients who didn't require readmission. On the nongeriatric ward, patients who were readmitted had a higher CGA completeness score than those who weren't readmitted. Even though this difference wasn't statistically significant, it's possible that on the geriatric ward, the more complete a CGA is, the more likely it is that a readmission event can be avoided. This result could be explained by the fact that, on the geriatric ward, a predefined CGA intervention package was implemented that included multidimensional assessment, multidisciplinary specialty expertise, geriatric meetings, care plans focused on patient-centered goals based on CGA results, the implementation of care plans by geriatric nurse specialists, and a continuous review of progress and care planning.4 In-hospital death. The rate of in-hospital death on the geriatric ward was about five times greater than that on the nongeriatric ward. Considering that the CGA was applied on the geriatric ward in the hospital analyzed in this study, this result was quite surprising. Nevertheless, the characteristics of patients on the geriatric ward—for example, being older and having more diseases—could be the reason why the mortality was higher. These results are also in line with a previous study, which showed that more older adults died on the geriatric ward than on the general ward.7 The relationship between the in-hospital mortality and the completeness of the CGA wasn't statistically significant. However, the mean CGA completeness score for older adults who didn't die in the hospital tended to be higher than that for patients who did, for both wards. This result is consistent with previous research, which stated that CGA can reduce mortality among hospitalized older adults.3,4,6 Length of stay. Length of hospital stay wasn't significantly related to age but was significantly related to comorbidities and ward type. The more illnesses a patient had, the longer their hospital stay. Further, older adults on the geriatric ward with more illnesses also stayed in the hospital longer than did those on the nongeriatric ward. This aligns with a previous study indicating that prolonged hospital stays often occur in older adults because of the higher incidence of complications, such as multimorbidity.9 The completeness of the CGA had a significant relationship with length of stay. For the total sample and on the geriatric ward, the relationship was positive and moderate or typical. This means that the more complete the CGA was, the longer the patient stayed in the hospital. This positive relationship is supported by previous studies.5-7 Limitations This study only evaluated the assessment part of the CGA, although there are three steps in the CGA process and six key features of CGA. The assessment part is the first stage. For effective implementation to improve care outcomes, the entire process and all of the key features need to be completed, although this wasn't the subject of the present study. Second, this study excluded older adults who were transferred to the ICU, High Care Unit, and Intensive Coronory Care Unit while they were hospitalized. This exclusion criterion was designed to control the condition of the older adults in both wards, so they were similar in acuity. However, this means the results underestimate the possibility that the older adults could die in the hospital, thus influencing mortality data. Third, the data for readmission were measured only by counting the rate of participants readmitted to the same hospital. Investigators didn't collect the reason for the readmission nor account for the possibility of patients being readmitted to another hospital. This could underestimate the readmission rate of the older patients. Fourth, the independent variables were few: age, comorbidity, ward type, and the completeness of CGA. The complex care outcomes of older adults might be influenced by many other factors, such as healthcare and patient factors, that warrant further investigation. Finally, these study results can only be generalized to similar settings and populations. Further revision of the 21-item evaluation for the CGA should be investigated in future studies. Implications for nurse leaders Nurse leaders, especially in geriatric settings, should understand that the CGA may not be optimally implemented in a hospital setting, and note which items aren't frequently assessed. The domains of functional status and psychological health in particular had a low completion rate. Although the entire team is responsible for addressing this concern, the nurse leader must identify and educate those who don't regularly complete full and complete evaluations. Involving nurses in the completion of the CGA would lead to more comprehensive assessments that ultimately improve care outcomes for older adults. These authors also challenge nurse leaders to get involved with policymaking to provide regulations for all medical teams that treat older adults, particularly in relation to geriatric assessments including the CGA. Getting to 100% The average CGA completeness score in this study was 68.19%, and the completeness rate in the geriatric ward was significantly higher than in the nongeriatric ward. There were three items that weren't documented in this study population: balance in the physical domain, mobility in the functional domain, and delirium in the psychological domain; these items should be emphasized for staff because they're important parameters in assessing the condition of hospitalized older adults. Of the four CGA domains, psychological health had the lowest average completeness rate, whereas socioenvironmental status had the highest average score. The overall evaluation showed that the CGA assessment wasn't optimally performed, even in the geriatric ward where it was the policy to do so. The 30-day readmission rate was 6.3%; more patients who were readmitted came from the nongeriatric ward compared with the geriatric ward, although this finding wasn't statistically significant. Similarly, the CGA completeness score of patients who weren't readmitted after 30 days was slightly higher than those who were readmitted, but again, this finding didn't reach statistical significance. In-hospital mortality was 8.1%, and older adults in the geriatric ward had significantly more deaths than those in the nongeriatric ward. However, those patients in both wards who survived had a more complete CGA. The median length of stay was 8 days (range, 2 to 62 days) and the length of stay in the geriatric ward was significantly longer than in the nongeriatric ward. Ultimately, the more complete the CGA, the longer the patient's length of stay.