Post-hospitalization mortality in the elderly

Post-hospitalization mortality in the elderly

Arch. Gerontol. Geriatr. 36 (2003) 83 /91 www.elsevier.com/locate/archger Post-hospitalization mortality in the elderly Maria Ponzetto *, Mauro Zano...

122KB Sizes 0 Downloads 74 Views

Arch. Gerontol. Geriatr. 36 (2003) 83 /91 www.elsevier.com/locate/archger

Post-hospitalization mortality in the elderly Maria Ponzetto *, Mauro Zanocchi, Barbara Maero, Erica Giona, Federica Francisetti, Elena Nicola, Fabrizio Fabris Institute of Gerontology, University of Torino, Corso Bramante 88, 10126 Torino, Italy Received 30 January 2002; received in revised form 29 July 2002; accepted 31 July 2002

Abstract The level of disability and polypathology in hospitalized elderly is usually high. Multidimensional and functional assessment allows to identify risk factors for clinical and functional failure of patients. Many studies point out that identifying predictors of highrisk patients is a necessary step in accurate targeting. We evaluated 395 subjects (175 women, 202 men, mean age 77.9 year) during their hospitalization in our Geriatric ward. Baseline data included: demographics variables, medical diagnosis, functional evaluation, and laboratory values. After a 6-month follow up 80 (20.2%) subjects died. In our study, male gender, dependence at the Dependence Medical Index (DMI), low serum albumin ( B/2.8 g/dl), impaired score at the Instrumental Activities of Daily Living scale (IADL), score lower than 13.7 at the acute physiology and chronic health evaluation (APACHE II) and neoplasm were independent predictors of 6-month post-hospitalization mortality. The high mortality rate of our sample could be a marker of considerable frailty among elderly patients. Our study shows that a poor functional status is a more reliable prognostic factor than type and number of admitting diagnosis. Clinical evaluation, improved with information about functional status, is a feasible and practical way of detecting risk of short term post-hospitalization mortality of elderly subjects. # 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Post-hospital mortality; Functional status of elderly

* Corresponding author. Tel.: /39-011-6637140; fax: /39-011-69611045 E-mail address: [email protected] (B. Maero). 0167-4943/02/$ - see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 1 6 7 - 4 9 4 3 ( 0 2 ) 0 0 0 6 1 - 4

84

M. Ponzetto et al. / Arch. Gerontol. Geriatr. 36 (2003) 83 /91

1. Introduction Hospitalized elderly have usually a high level of disability and polypathology (Poli et al., 1993). Usually an acute medical illness requiring hospitalization is followed by a progressive physical decline, resulting in high rates of mortality during months following discharge. Prognostic information collected during the hospitalization improves the definition of the goals of care and therapy. Multidimensional assessment in these patients is useful to identify acute and chronic pathological events and their implication on functional status. Moreover it points out social and economical problems, which can interact with disabling consequences of illnesses (Cohen et al., 1992; Kane, 1993; Thomas et al., 1993). Many studies show that in-hospital mortality (Narain et al., 1988; Pompei et al., 1991; Winograd et al., 1991), risk of institutionalization (Jagger et al., 1993) and post-hospitalization mortality (Inouye et al., 1998) are closely linked to patient functional status. Impaired cognitive function has been implicated in immediate mortality, prolonged stay, and nursing home placement. The aim of this study is to evaluate 6-month mortality and its predictors among a cohort of elderly subjects admitted to the University Geriatric ward in Torino, Italy.

2. Subjects and methods Data were collected by clinical records used in our Geriatric ward (Fabris et al., 1989). Baseline data included: demographics variables, medical diagnoses, functional evaluation, and laboratory values. Potential participants were 500 patients admitted to the University Department of Geriatric Medicine in Torino, between August 1, 1999 and April 1, 2000. Fifty-two patients were excluded because they died during hospitalization. Fifty-three subjects could not be interviewed at follow-up, because they changed address or telephone number. The final sample included 395 patients. They were followed-up, making a monthly telephone interview. Information about their vital status and utilization of hospital and nursing home care was collected. Death causes according to ICD-9-CM (The International Classification of Disease, 1989) were gathered from local statistical office death certificates. Demographic, clinic and functional status information obtained within 2 days of admission was considered as possible predictors of 6-month post-hospitalization mortality. Age, gender, marital status were considered as demographic data. We also take into account discharge interventions, i.e. home care, institutionalization. Instrumental Activities of Daily Living scale (IADL) (Lawton and Brody, 1969) and index of Activities of Daily Living (ADL) (Katz et al., 1970) were used to assess functional status. Cognitive status was evaluated by the Short Portable Mental Status Questionnaire (SPMSQ) (Pfeiffer, 1975). Patients were classified as being independent or dependent for medical reasons using the Dependence Medical Index (DMI) (Fabris et al., 1996). The DMI groups seven major criteria (severe

M. Ponzetto et al. / Arch. Gerontol. Geriatr. 36 (2003) 83 /91

85

impairment in strength and/or motility in at least two limbs; double incontinence; pressure sores; severe disturbances in speech and communication; severe decline in sight and/or hearing not modifiable with prosthesis; terminal illness (death expected within 3/6 months); need for multiple and complex therapies) and eight minor criteria (mild loss of strength and/or motility in at least two limbs; unstable incontinence, episodic disorientation; moderate disturbances in speech and communication, decline in sight and/or hearing only partially modifiable with prosthesis; dizziness with tendency to fall; chronic relevant diseases in unstable balance; permanent but independent use of the wheel-chair). The presence of at least three minor criteria is a determinant of dependence for medical reasons, while only one major criterion was needed. The number of pathologies was evaluated. The following primary admitting diagnosis categories were considered: cardiac, cerebrovascular, pulmonary, gastrointestinal, muscle-skeleton, dementia, neoplastic or other. Patient’s clinical severity was ascertained using the acute physiology and chronic health evaluation (APACHE II). The cut point of more than 16 usually indicates high burden illness (Knaus et al., 1985). In our sample none had APACHE II index’s score higher than 16; thus we considered the 95th percentile, 13.7 APACHE II score. Electrocardiogram (ECG) was performed to point out presence of: atrial fibrillation, left ventricular hypertrophy or ischemic modifications. The body mass index (BMI), systolic (SBP) and diastolic (DBP) blood pressure, heart rate, serum value of albumin, hemoglobin, total cholesterol, creatinine, sodium, potassium evaluated 3 days before discharge were considered and used for statisical elaboration. The number of medications prescribed at discharge was also recorded. Distribution curves were obtained for continuing variables; we calculated distribution curves at the 5th percentile for hemoglobin, albumin, total cholesterol, creatinine, sodium, potassium, DBP, SBP, and at the 95th percentile for heart rate, age, number of prescribed drugs, number of pathology and days of hospitalization. The x2-test was used to assess univariate association between mortality and categorical variables; variance analysis was chosen for continuous variables. Multivariable Cox analysis was also used. We obtained survival curves by Kaplan Mayer’s model.

3. Results Sample population included 395 subjects, monitored by telephone at 6-month after discharge. Mean age of the study population was 79 years; 55.7% were male 40.2% were widowed and 19.7% were living alone. Complete baseline characteristics of the sample were reported in Table 1. At the end of follow-up period 23 subjects were institutionalized (7 male and 16 female), 292 were living at home (157 male and 135 female). Eighty subjects (56 male and 24 female) died during the 6-month followup; 70 were at home. Among them 10 subjects were attended by the Home Care Service of our Geriatric Department. Men were at higher risk of dying than women (25.5% men against 13.7% women, with a relative risk (RR) of 2.15).

86

M. Ponzetto et al. / Arch. Gerontol. Geriatr. 36 (2003) 83 /91

Table 1 Clinical and functional characteristics of 395 patients n (%) Male DMI (dependency) ECG alterations Ischemic modifications Left ventricular hypertrophy Atrial fibrillation Primary admitting diagnosis category Cardiovascular disease Cerebrovascular disease Pulmonary disease Gastrointestinal disease Muscle-skeleton disease Dementia Neoplasm Other conditions

220 (55.7) 90 (22.8) 31 (7.8) 13 (3.3) 81 (20.5) 56 (14.1) 44 (11.1) 40 (10.1) 43 (10.9) 18 (4.5) 26 (6.5) 46 (11.6) 122 (30.8) Mean9/SD

Age APACHE score BMI SBP (mmHg) DBP (mmHg) Heart rate (pulse/min) Albumin (g/dl) Hemoglobin (g/dl) Total cholesterol (mg/dl) Creatinine (mg/dl) Sodium (mEq/l) Potassium (mEq/l) Length of stay (days) Prescription medications SPMSQ (no. of mistakes) IADL (score) ADL (no. of lost function)

77.99/8.2 9.359/2.3 23.159/5.6 1499/23.2 84.359/10.6 82.759/12.9 3.739/0.6 12.129/2.4 175.439/44.3 1.449/1.18 139.829/4.5 4.089/0.6 17.229/22.9 3.479/2.1 2.709/2.9 7.609/5.1 1.579/2.2

Dead subjects were, at discharge, older (mean9/SD age of 88.49/8.2 years) than survivors (mean9/SD age of 77.29/8.5 years) (P B/0.05). Marital status was not significantly associated with mortality. At discharge, 18 subjects returned to their home followed by our Service of home care, 55.6% of this sample died during the 6-month follow up, and 44.4% survived (RR /5.48). Six-months survival was associated with a better IADL level (P B/0.001). Dead subjects had a number of ADL impairments significantly higher than survivors (P B/ 0.001). Cognitive status was worse in dead subjects than in alive patients (P /0.002)

M. Ponzetto et al. / Arch. Gerontol. Geriatr. 36 (2003) 83 /91

87

Twenty-nine DMI dependent subjects died (32.2%), against 16.7% of DMI independent patients (RR /2.37). Among DMI absolute criteria only the presence of terminal illness was significantly associated with mortality. Hospital length of stay and number of illnesses were not significantly different between the two groups. During the 6 months after discharge, dead subjects at follow-up had 0.199/0.19 hospital re-admissions and survivors’ re-hospitalization rate was 0.379/0.83 (P /0.004). None of the principal discharge diagnosis, except neoplasm, was associated with mortality. Dead subjects had a significantly higher APACHE II score (P /0.04). BMI, blood pressure, heart rate and electrocardiographic alterations were not significantly associated with mortality. Dead subjects had lower albumin, hemoglobin and sodium values than survivors. Blood potassium was higher in dead patients than survivors. There were no significant differences between the two groups for creatinine and total cholesterol values. The number of medications prescribed at discharge was, respectively 3.569/2.33, for dead subjects at follow-up, and 3.449/1.39, for survivors (P /0.04) (Table 2). Risk factors for 6-month mortality after discharge in multivariate analysis were: male sex, albumin 5/2.8 g/dl, diagnosis of neoplasm, DMI and IADL dependence, APACHE II score /13.7 (Table 3).

4. Discussion Comprehensive geriatric assessment is useful to point out peculiar risk factors for mortality in the elderly. ADL and IADL dependence, social isolation, poor physical activity, and living in bad economic conditions are associated with a higher risk of death (Scott et al., 1997). After discharge, one-third of the over 75-year-old subjects had shown one more new disability (Mayer-Oakes et al., 1991). Our study shows a percentage of 20.2 of mortality at the end of the 6-month follow-up. According to other published studies, high mortality is a sign of frailty of elderly people’s health. Dead subjects during follow-up are significantly older (mean age about 90 years), than survivors (mean age about 80 years). Advanced age is related to poor functional status and death. When some illness appears, older subjects are at higher risk of clinical and functional failure and often of death. Nevertheless our multivariate analysis does not point out a clear association between older age and mortality. In accordance with other studies, we find that older age becomes a less important predictive variable if others, like severity of admitting diagnosis, presence of chronic illness and poor functional status, are taken into account (Mayer-Oakes et al., 1991). The fact that after discharge there were more subjects followed at home by our specific home care service (nurses and doctors from hospital to patient’s house)

88

M. Ponzetto et al. / Arch. Gerontol. Geriatr. 36 (2003) 83 /91

Table 2 Risk factors associated with 6-month mortality: univariate analysis Risk factors (mean9/SD)

Dead

Alive

P value*

Age N. pathology APACHE score BMI SBP (mmHg) DBP (mmHg) Heart rate (pulse/min) Albumin (g/dl) Hemoglobin (g/dl) Total cholesterol (mg/dl) Creatinine (mg/dl) Sodium (mEq/l) Potassium (mEq/l) Lenght of stay (days) Prescription medications SPMSQ (no. of mistakes) IADL (score) ADL (no. lost function) Rehospitalization

88.49/8.2 2.609/1.27 11.119/1.81 22.359/6.22 144.449/22.96 82.629/9.28 85.109/12.87 3.539/0.60 11.689/2.72 169.319/43.72 1.479/1.06 138.819/6.09 4.129/0.65 16.399/18.19 3.569/2.33 3.459/3.29 5.209/4.67 2.599/2.48 0.199/0.19

77.29/8.5 2.679/1.3 8.909/2.2 22.359/6.2 150.169/23.2 84.799/10.9 82.159/12.9 3.779/0.50 12.249/2.30 176.999/44.3 1.449/1.2 140.079/3.96 4.079/0.56 17.449/24.0 3.449/1.99 2.519/2.78 8.219/5.0 1.319/2.01 0.379/0.83

0.04 0.69 0.04 0.07 0.66 0.15 0.39 0.02 0.03 0.74 0.73 0.001 0.03 0.97 0.04 0.002 B/0.001 B/0.001 0.004 RR (95% CI)a

Male (n , %) DMI (dependency) (n , %)

56 (25.4%) 29 (32.2%)

164 (74.5%) 61 (67.8%)

2.15 (1.40 /3.55) 2.37 (1.39 /4.04)

ECG alterations (n , %) Ischemic modifications Left ventricular hypertrophy Atrial fibrillation

7 (22.6%) 3 (23.1%) 20 (24.7%)

24 (77.4%) 10 (76.9%) 61 (75.3%)

1.48 (0.88 /1.07) 1.93 (0.90 /2.30) 1.38 (0.77 /2.47)

48 (85.7%) 37 (84.1%) 32 (80.0%) 35 (81.4%) 17 (94.6%) 22 (84.6%) 22 (47.8%) 102 (83.6%)

0.62 0.72 0.98 0.89 0.22 0.76 5.71 0.70

Primary admitting diagnosis category (n , %) Cardiovascular disease 8 (14.3%) Cerebrovascular disease 7 (15.9%) Pulmonary disease 8 (20%) Gastrointestinal disease 8 (18.6%) Muscle-skeleton disease 1 (5.6%) Dementia 4 (15.4%) Neoplasm 24 (52.2%) Other conditions 20 (16.4%)

(0.28 /1.36) (0.31 /1.68) (0.43 /2.22) (0.39 /2.00) (0.29 /1.69) (0.23 /2.09) (3.00 /10.90) (0.40 /1.21)

* Variance analysis. a x2-test analysis.

among dead people (10/80, 12.5%) than survivors (8/315) (2.5%) means that an early and protected discharge could be also carried out for extremely critically ill subjects. The number of ADLs needing physical assistance is significantly higher between dead subjects than survivors. IADL score is significantly lower for the first group. IADL is an independent predictor of post-hospitalization mortality (Fig. 1).

M. Ponzetto et al. / Arch. Gerontol. Geriatr. 36 (2003) 83 /91

89

Table 3 Risk factors associated with 6-month mortality: multivariate analysis

Male gender DMI dependence Albumin B/2.8 g/dl IADL dependence Neoplasm APACHE score /13.7

b

E.S. (b)

RR

95% I.C.

P

0.84 0.82 0.87 0.64 1.22 0.37

0.29 0.28 0.37 0.32 0.27 0.07

2.31 2.28 2.39 1.89 3.40 1.45

1.31 /4.07 1.31 /3.99 1.16 /4.89 1.01 /3.56 2.01 /4.91 1.27 /1.65

B/0.01 B/0.01 B/0.05 B/0.05 B/0.00001 B/0.00001

Our study does not differ from many others showing that a poor functional status is a more reliable prognostic factor than type and number of admitting diagnosis. Impaired functional status is due not only to admitting diagnosis but also to other health problems unknown at hospital admission.

Fig. 1. Cumulative survival related to IADL.

90

M. Ponzetto et al. / Arch. Gerontol. Geriatr. 36 (2003) 83 /91

Dead subjects at the 6-month follow-up had a significantly higher SPMSQ score than survivors; Cohen showed similar findings among a sample of people older than 75 (Cohen et al., 1992). In our sample low mental status is not independently associated with mortality at multivariate analysis. Mortality rate is 32.2% for DMI dependent subjects and 16.7% for DMI independent patients. Dependence due to clinical reasons is an independent risk factor of post-hospital mortality. Among DMI absolute criteria, only terminal illness, ascertained by medical staff, is significantly associated with mortality. An accurate evaluation of different clinical disability factors is necessary to improve the definition of prognosis. In our study neoplasm is an independent predictor of short-term mortality after discharge. Albumin value B/2.8 g/dl is an independent risk factor of posthospitalization mortality. Low albumin values are related to the type of disease and its severity, in particular they are tightly associated with malnutrition, neoplasm and polypathology (Covinsky et al., 1999). We found that the risk of 6-month post-hospitalization mortality is higher in male subjects. In many studies men have higher mortality rate than women, independent of their age. Women are probably more protected because they have a lower hormon-dependent risk of cardiovascular diseases and a better immunitary system than men (Jagger et al., 1993). Clinical evaluation, improved with information about functional status, is a feasible and practical way of detecting risk of short term posthospitalization mortality of elderly subjects and of recognizing frail subjects and their needs of care. In our opinion it is important to consider clinical and functional risk factors for short term post-hospitalization mortality, in order to plan a cautious discharge and periodic medical/nursing controls of frail patients.

References Cohen, H.J., Saltz, C.C., Samsa, G., McVey, L.J., Davis, D., Feussner, J.R., 1992. Predictors of two-year post-hospitalization mortality among elderly veterans in a study evaluating a geriatric consultation team. J. Am. Geriatr. Soc. 40, 1231 /1235. Covinsky, K.E., Martin, G.E., Beyth, R.J., Justice, A.C., Sehgal, A.R., Landfeld, C.S., 1999. The relationship between clinical assessment of nutritional status and adverse outcomes in older hospitalized medical patients. J. Am. Geriatr. Soc. 47, 532 /538. Fabris, F., Macchione, C., Molaschi, M., Ferrario, E., Pernigotti, L., Visentin, P., 1989. La cartella clinica geriatrica. Una proposta di valutazione multidimensionale. Minerva Med. 80 (Suppl. 12), 1 /56(in Italian). Fabris, F., et al., 1996. Dependence Medical Index (DMI) in elderly persons: a tool for identification of dependence for medical reasons. Bold 2, 9 /14. Inouye, S.K., Peduzzi, P.N., Robison, J.T., Hughes, J.S., Horwitz, R.I., Concato, J., 1998. Importance of functional measures in predicting mortality among older hospitalized patients. J. Am. Med. Assoc. 279, 1187 /1193. Jagger, C., Spiers, N.A., Clarke, M., 1993. Factors associated with decline in function, institutionalization and mortality of elderly people. Age Ageing 22, 190 /197. Kane, R.L., 1993. The implications of assessment. J. Gerontol. (spec. issue) 48, 27 /31.

M. Ponzetto et al. / Arch. Gerontol. Geriatr. 36 (2003) 83 /91

91

Katz, S., Downs, T.D., Cash, H.R., Grotz, R.C., 1970. Progress in development of the index of ADL. Gerontologist 10, 20 /30. Knaus, W.A., Draper, E.A., Wagner, D.P., Zimmerman, J.E., 1985. APACHE II: a severity of disease classification system. Crit. Care Med. 13, 819 /829. Lawton, M.P., Brody, E.M., 1969. Assessment of older people: self-maintaining and instrumental activities of daily living. Gerontologist 9, 179 /186. Mayer-Oakes, S.A., Oye, R.K., Leake, B., 1991. Predictors of mortality in older patients following medical intensive care: the importance of functional status. J. Am. Geriatr. Soc. 39, 862 /868. Narain, P., Rubenstein, L.Z., Wieland, G.D., Rosbrook, B., Strome, L.S., Pietruszka, F., Morley, J.E., 1988. Predictors of immediate and 6-month outcomes in hospitalized elderly patients. The importance of functional status. J. Am. Geriatr. Soc. 36, 775 /783. Pfeiffer, E., 1975. A short portable mental status questionnaire for the assessment of organic brain deficit in elderly patients. J. Am. Geriatr. Soc. 23, 433 /441. Poli, L., Pich, A., Zanocchi, M., Fonte, G., Bo, M., Fabris, F., 1993. Autopsy and multiple patology in the elderly. Gerontology 39, 55 /63. Pompei, P., Charlson, M.E., Ales, K., MacKenzie, C.R., Norton, M., 1991. Relating patient characteristics at the time of admission to outcomes of hospitalization. J. Clin. Epidemiol. 44, 1063 /1069. Scott, W.K., Macera, C.A., Cornman, C.B., Sharpe, P.A., 1997. Functional health status as a predictor of mortality in men and women over 65. J. Clin. Epidemiol. 50, 291 /296. The International Classification of Disease, 9th revision, Clinical Modification: ICD-9-CM. 3rd ed., 1989. US Department of Health and Human Services, US Public Health Services Publication (PHS) 89, 1260. Washington DC. Thomas, D.R., Brahon, R., Haywood, B.P., 1993. Inpatient community-based geriatric assessment reduces subsequent mortality. J. Am. Geriatr. Soc. 41, 101 /104. Winograd, C.H., Gerety, M.B., Chung, M., Goldstein, M.K., Dominguez, F., Jr., Vallone, R., 1991. Screening for frailty: criteria and predictors of outcomes. J. Am. Geriatr. Soc. 39, 778 /784.