Comorbidity and intercurrent diseases in geriatric stroke rehabilitation: a multicentre observational study in skilled nursing facilities

Comorbidity and intercurrent diseases in geriatric stroke rehabilitation: a multicentre... Background Older patients often have multiple comorbidities and are susceptible to develop intercurrent diseases during rehabilitation. This study investigates intercurrent diseases and associated factors in patients undergoing geriatric stroke rehabilitation, focussing on pre-existing comorbid conditions, overall comorbidity and baseline functional status. Materials and methods This multicentre prospective cohort study included 15 skilled nursing facilities. Data were collected at baseline and at discharge. The primary outcome measures were presence and number of intercurrent diseases. Further- more, their impact on change in rehabilitation goals or length of stay was examined. Comorbidity was assessed with the Charlson index, and functional status with the Barthel index (BI). Results Of the 175 included patients, 51% developed an intercurrent disease. A lower baseline BI, a higher Charlson index, presence of diabetes mellitus (DM) and kidney disease were related to the occurrence of an intercurrent disease (p < 0.05). Moreover, a lower BI, a higher Charlson index, and particularly the presence of DM were independently associated. If both comorbidity and a lower baseline functional status were present, the odds ratio (95% CI) of developing intercurrent diseases was 6.70 [2.33–19.2], compared to 1.73 [0.52–5.72] (comorbidity only) and 1.62 [0.53–4.94] (only BI ≤ 14). Conclusions On admission, functional impairments and comorbidity, particularly diabetes, independently contribute to developing intercurrent diseases during geriatric stroke rehabilitation. Therefore, routine evaluation of comorbidity integrated with functional status at the start of rehabilitation is essential to identify patients at risk. Finally, particular attention should be paid to patients with DM to prevent intercurrent diseases and support optimal functional recovery . Keywords Geriatric rehabilitation · Comorbidity · Functional impairments · Intercurrent diseases · Stroke Introduction Electronic supplementary material The online version of this article (https ://doi.org/10.1007/s4199 9-018-0043-5) contains supplementary material, which is available to authorized users. Following acute hospitalisation, rehabilitation helps patients to regain functional independency that enables them to be * Anouk D. Kabboord discharged home. However, during hospitalisation, the A.D.Kabboord@lumc.nl risk of functional decline and complications is particularly Department of Public Health and Primary Care, Leiden increased in older patients [1]. In the Netherlands, about University Medical Center, Hippocratespad 21, Postbus one-third of all stroke patients are referred to a skilled nurs- 9600, 2300 RC Leiden, The Netherlands ing facility (SNF) that provides geriatric rehabilitation. Department of Neurology, Erasmus MC University Medical These patients are usually relatively older, have a longer Center, ‘s-Gravendijkwal 230, Rotterdam, The Netherlands length of stay (LoS) in the acute hospital, and have more Department of Primary and Community Care, Centre complex problems (Supplement material Appendix A) [2]. for Family Medicine, Geriatric Care and Public Health, Also, during inpatient rehabilitation, intercurrent diseases Radboud University Nijmegen Medical Centre, Nijmegen, may occur that interfere with therapy and could negatively The Netherlands 4 impact rehabilitation outcome [3, 4]. “Joachim en Anna”, Centre for Specialized Geriatric Care, Nijmegen, The Netherlands Vol.:(0123456789) 1 3 348 European Geriatric Medicine (2018) 9:347–353 Studies investigating complications during inpatient prolonged the LoS or (ii) whether the rehabilitation goals stroke rehabilitation found that 30–96% of the patients needed adjustment. Four categories were formed: (1) no developed complications; this wide range could be due to intercurrent disease, (2) ‘No impact’, (3) ‘With impact’, and different definitions of a complication and the methods of (4) intercurrent disease that directly caused death. measurement [5–12]. The present study investigates inter- current diseases, i.e. any disease that occurs during the Data collection progress of another disease, during rehabilitation. Factors related to intercurrent diseases can include age, gender [9, The participating multidisciplinary teams consisted of a 13], time interval between stroke and rehabilitation [7, 10, physician [16], a physiotherapist, an occupational therapist, 12], severe stroke [7, 11] or functional impairment [6, 9, 12, a psychologist, a speech therapist, a dietician and skilled 13] and comorbidity [5–7, 13, 14], although it is unknown nurses; all received the same instructions regarding perfor- which specific comorbidities are related. Particularly, older mance of the assessments. Data were collected within the patients are at risk of functional decline and often have mul- first 2 weeks after admission (T0) and at discharge (T1) from tiple comorbidities. However, few studies have investigated the SNF or (at the latest) 1 year after admission, if a patient associations with intercurrent diseases in the older, vulner- was still in the SNF at that time. able group of patients receiving geriatric stroke rehabilita- tion [5, 13]. Furthermore, intercurrent diseases may impede Measurements successful functional recovery [5, 14]. Therefore, to better understand the relations between comorbidity, functional The following patient characteristics and data were col- impairment and intercurrent diseases, and to identify asso- lected: age, gender, home situation, comorbidity, LoS in ciated pre-existing comorbid conditions, this study explores: acute hospital, LoS in the SNF, and discharge destination (i) the presence, and number of intercurrent diseases and [5–14, 17]. Functional assessment was performed at base- their impact on older patients admitted to an SNF, recover- line and at discharge using the modified Barthel index (BI) ing after stroke, and (ii) factors associated with the presence to assess activities of daily living (ADL) [18]. Premorbid and number of intercurrent diseases, focusing on functional BI was assessed on admission, using information on the status and comorbidity. patient’s situation prior to the acute stroke, based on inter- view and collateral history. Functional recovery was defined in two ways: BI at discharge and ‘relative functional gain’, Materials and methods which was calculated as follows: (BI-discharge minus BI- admission)/(BI-premorbid minus BI-admission) × 100 [19, Participants 20]. Relative functional gain expresses the achieved percent- age of potential functional gain. Data were obtained from the Geriatric Rehabilitation in Pre-existing comorbidity was assessed using the Charl- AMPutation and Stroke (GRAMPS) study. Data collection son comorbidity index (Charlson-CI). This index consists took place between January 2008 and July 2010; details of 19 diagnoses and was adjusted for stroke [21–23]. The on the study design are already published [15]. A total of Charlson-CI was categorised as: 0 (no comorbidity), 1 (sin- 15 SNFs located in the southern part of the Netherlands gle comorbidity) or ≥ 2 (multiple comorbidities), unless oth- participated. All stroke patients admitted to one of these erwise specified. Comorbidities were recorded if present in SNFs were eligible for inclusion. Patients were excluded medical history, e.g. chronic diseases and conditions that if they refused participation, were unable to give informed required ongoing use of (preventive) medication. Conditions consent, were critically ill, or were expected to have a stay that had completely resolved without any residual symp- of ≤ 2 weeks. The medical ethics committee of the region toms or need for treatment were not noted (e.g. childhood Nijmegen-Arnhem approved the study protocol. asthma). Finally, if myocardial infarction in the past had led to heart failure, only heart failure was recorded. Outcome measures Statistical analysis For the present study, the outcome measures were: the pres- ence and number of intercurrent diseases that occurred dur- Data were processed and analysed using the Statistical Pack- ing rehabilitation. Intercurrent diseases were coded using the 10th revision Clinical Modification ICD-10CM. At dis- age for Social Science version 23. Means with standard deviations (normal distribution), medians with interquartile charge, the attending physician registered intercurrent dis- eases that affected the course of the rehabilitation: impact ranges (skewed data), or absolute numbers with percentages (categorical data) are reported. was classified according to (i) whether the disease had 1 3 European Geriatric Medicine (2018) 9:347–353 349 A Chi-Squared test (categorical data), ANOVA or Characteristics related to intercurrent diseases Kruskal–Wallis test, depending on their distribution, were used to detect mean differences in characteristics Patients without any intercurrent disease had a BI on admis- between the four intercurrent disease categories and to sion of at least 4 points higher than those with intercurrent identify comorbid conditions related to the occurrence of diseases. The proportion of patients without comorbidity was intercurrent diseases. A p value of ≤ 0.05 was considered largest in the category ‘no intercurrent disease’ (52%), whereas statistically significant. in the category ‘With impact’, the proportion of patients with Multivariate analyses were performed using binary multiple comorbidities was the largest (54%), p = 0.007. logistic regression with the presence of intercurrent dis- Patients that developed intercurrent diseases were less often eases and Poisson regression with number of intercur- discharged home, had a longer LoS, a lower BI at discharge, rent diseases as the dependent variable. Rehabilitation and a lower relative functional gain. This also applied to the LoS (log) was added as the ‘offset’. Factors included in category that was considered as having ‘No impact’. Mul- the multivariate model were age and gender. Significant tivariate analyses showed that: BI on admission (OR 0.87 baseline variables (p < 0.10) were added as a continuous [0.82–0.92]) and comorbidity (OR 1.43 [1.13–1.81]) were variable if applicable. independently associated with the presence of intercurrent Before performing the analyses, data were tested for diseases, but only the Charlson-CI was significantly associated the required assumptions, such as multicollinearity, inter- with number of intercurrent diseases (incidence rate ratio: 1.14 action and effect modification. To investigate comorbid- [1.03–1.25], p: 0.008). This means that with every extra point ity and baseline functional status, separate and combined on the Charlson-CI, a 14% increase in the number of intercur- relations with the presence of intercurrent diseases were rent diseases is expected (Supplement material Appendix D). analysed. For this purpose, variables were dichotomized. Odds ratios (OR) were calculated with the absence of both Comorbidity and intercurrent diseases factors as reference category [24]. Sensitivity analyses were performed, i.e. with and without deceased patients. Having diabetes and/or kidney disease was significantly related to the occurrence of an intercurrent disease (Table 2). More- over, logistic regression analysis showed that only diabetes was independently associated (OR 3.50 [1.32–9.26]). No clear Results patterns or relations between comorbidities and specific inter - current diseases were observed: a wide variety of different Characteristics diseases occurred in patients with pre-existing comorbidity. The intercurrent diseases that most frequently occurred were Of the 378 eligible patients, 186 were included in the cardiovascular (13%), psychiatric (12%) such as depression GRAMPS study; the excluded patients did not differ and delirium, and genitourinary (11%), predominantly uri- with regard to age, gender or LoS [25]. The present study nary tract infections. An overview of intercurrent diseases, included 175 patients because 11 patients were lost to per comorbidity (the five most prevalent only), is presented in follow-up, mainly due to translocation to another SNF Supplement material Appendix E. (Supplement material Appendices B and C). Table 1 pre- sents the baseline characteristics of the study population, Comorbidity and baseline functional status and the intercurrent disease categories. Mean age was 78.8 years and 46% were males. On average, LoS in the Table 3 shows the cumulative effect of the combination of a acute hospital was 19  days, the premorbid BI was 20, lower functional status on admission (BI ≤ 14) and the pres- baseline BI was 12, and BI at discharge was 17. LoS in ence of comorbidity (Charlson-CI ≥ 1) in relation to the occur- the SNF was 12 weeks, the (average) relative functional rence of an intercurrent disease. On admission, when comor- gain was 67, and 56% of these patients was discharged bidity and lower functional status on admission were present home. separately, ORs were 1.73 [0.52–5.72] and 1.62 [0.53–4.94], Of the 89 (51%) patients that developed an intercurrent respectively. However, if both were present, the OR was 6.70 disease, 49% developed one disease, 33% ≥ 2 diseases, [2.33–19.2]. and 18% died. Comorbidity was present in 116 (62%) patients: 40 (21%) scored 1 and 76 (41%) scored ≥ 2. The most prevalent pre-existing comorbidities were myocar- dial infarction (18%), diabetes mellitus (18%) and conges- tive heart failure (16%). 1 3 350 European Geriatric Medicine (2018) 9:347–353 Table 1 Patient characteristics classified by intercurrent disease (ID) impact category Total baseline ID absent ID no impact ID with impact ID deceased ID impact n = 175 n = 86 n = 22 n = 46 n = 16 unknown n = 5 Variables at baseline Age (years), Mean 78.8 (8.0) 78.2 (8.3) 78.8 (5.6) 78.9 (8.5) 81.2 (8.4) 82.6 (7.8) (SD) Gender (male), n (%) 80 (46) 45 (52) 11 (50) 16 (35) 7 (44) 1 (20) Charlson-CI score, 1 (2)* 0 (2) 1 (2) 2 (2) 2 (2) 2 (3) median (IQR) Charlson-CI = 0, n 68 (39) 45 (52) 8 (36) 10 (22) 3 (19) 2 (40) (%) Charlson-CI = 1, n 38 (22) 19 (22) 6 (27) 11 (24) 2 (13) 0 (0) (%) Charlson-CI ≥ 2, n 69 (39) 22 (26) 8 (36) 25 (54) 11 (69) 3 (60) (%) Premorbid Barthel 20 (3) 20 (2) 20 (2) 19 (3) 17 (7) 18 (3) Index, median (IQR) LoS acute hospital in 19 (14) 19 (11) 19 (13) 19.5 (18) 22 (18) 21 (21) days, median (IQR) Barthel Index on 12 (10)* 14 (7) 9 (12) 9 (8) 8 (9) 10 (6) admission, median (IQR) Variables at discharge LoS rehabilitation 12 (15)* 8 (6) 16 (23) 22 (26) – 16 (6) in weeks, median (IQR) Barthel Index at 17 (8)* 18 (4) 16 (9) 11 (10) – 15 (4) discharge, median (IQR) Relative functional 67 (90)* 85 (84) 67 (76) 24 (79) – 71 (42) gain, median (IQR) Discharge home, n 88 (56) 62 (73) 9 (43) 13 (28) – 4 (80) (%) Equal statistical significance was found when deceased patients were excluded SD standard deviation, Charlson-CI Charlson comorbidity index, IQR interquartile range, LoS length of stay Statistical significance at p < 0.05 *Kruskal–Wallis test Chi-Square test functional status, higher pre-existing comorbidity burden in Discussion general and specifically the presence of diabetes mellitus were independent determinants of developing intercurrent Main findings diseases. Furthermore, patients with multiple comorbidi- ties (higher Charlson-CI) had an increased risk to develop a To our knowledge, this is the first study to focus on comor - higher number of intercurrent diseases. Finally, the odds of bidity and intercurrent diseases during geriatric stroke reha- developing an intercurrent disease were substantially higher bilitation. The study cohort was characterised by a large drop if a patient had both comorbidity and functional impairment in functional status after acute stroke, often with multiple than if only one of these factors were present. comorbidities and a higher age compared to the majority of studies on stroke patients [5–10, 12–14, 26]. Although this subgroup had been triaged for inpatient geriatric reha- Intercurrent diseases bilitation, and selected as a vulnerable subgroup of patients on the base of medical complexity and functional depend- The percentage of patients (51%) that developed inter- ency, discriminant factors were still present. Lower baseline current diseases is comparable to that of studies using an 1 3 European Geriatric Medicine (2018) 9:347–353 351 Table 2 Associations between Comorbidity Total ID absent ID present comorbid conditions and # # presence of ≥ 1 intercurrent Charlson index, median (IQR) 1 (2) 0 (2) 2 (3) disease (ID) Comorbid condition Myocardial infarction, n (%) 31 (18) 13 18 Heart failure, n (%) 29 (17) 10* 19* Peripheral vascular disease, n (%) 23 (13) 9 14 Dementia, n (%) 1 (1) 1 0 Chronic pulmonary disease, n (%) 18 (10) 8 10 Musculoskeletal/connective tissue, n (%) 9 (5) 2* 7* Ulcers, n (%) 8 (5) 2 6 Mild liver disease, n (%) 3 (2) 1 2 # # Kidney disease (moderate), n (%) 16 (9) 3 13 # # Diabetes mellitus, n (%) 31 (18) 9 22 Malignancy, n (%) 10 (6) 3 7 Leukaemia, n (%) 1 (1) 1 0 Lymphoma, n (%) 2 (1) 0 2 Moderate liver disease, n (%) 0 (0) 0 0 Metastasis of solid tumour, n (%) 3 (2) 1 2 Any malignancy (of the above mentioned), n (%) 13 (7) 4 9 Comorbidities included in the logistic regression analysis are presented in bold Chi Square test: * p < 0.10, p < 0.05 Table 3 Comorbidity and baseline function: the separate and com- rates were similar to those in studies using prospective bined effect on developing an intercurrent disease in geriatric stroke assessment and similar prevalent diseases were found, i.e. rehabilitation (n = 170) genitourinary (urinary tract infections) and psychiatric dis- Charlson-CI BI ≤ 14 on Intercurrent Odds ratio [95% CI] eases (depression and delirium) [6–12, 14, 17]. However, score ≥ 1 admission* disease in the present study intercurrent cardiovascular disease was more prevalent, presumably because pre-existing Yes No cardiovascular comorbidities were highly prevalent in our No No 6 17 Reference 1.00 subgroup of vulnerable geriatric patients. No Yes 16 28 1.62 [0.53–4.94] Yes No 11 18 1.73 [0.52–5.72] Yes Yes 52 22 6.70 [2.33–19.2] Intercurrent diseases and their associations Charlson-CI charlson comorbidity index, BI Barthel index; CI confi- dence interval The presence of intercurrent diseases was related to reha- *Assessing the BI on admission was not possible in 5 patients. Sen- bilitation impact indices (longer LoS, less functional sitivity analysis showed similar results: when deceased patients were recovery and less often being discharged home). Despite excluded (n = 154) ORs were 1.32 [0.42–4.11], 1.42 [0.42–4.83] and that physicians retrospectively registered intercurrent dis- 5.54 [1.91–16.0], respectively eases according to their influence on rehabilitation, it was striking that this relation also applied to the category ‘No assessment method similar to ours (i.e. 30–54%) [5, 8, impact’. This underlines the impact that intercurrent dis- eases can have on rehabilitation outcomes. Besides base- 9, 13, 14, 17]. However, although other studies found a higher rate (60–100%), there was a clear difference in the line functional status and comorbidity in general, diabetes mellitus was found to be a significant determinant of the methods used. For example, shoulder pain, limb spasticity, dysphagia or aphasia were categorised as a complication, occurrence of an intercurrent disease. Diabetes affects various organ systems (e.g. vascular, skin, eyes, nervous whereas in the present study (and similar studies) these were considered to be symptoms and not diseases [6, 7, system) and might be the (underlying) cause of a variety of intercurrent diseases. However, the present study had 10–12, 26]. In this study, we were specifically interested in intercurrent diseases that occurred during the inpatient insufficient power to further investigate different comor - bidities and their associations with specific intercurrent rehabilitation period, and physicians retrospectively regis- tered the intercurrent diseases. Nevertheless, our incidence diseases. 1 3 352 European Geriatric Medicine (2018) 9:347–353 some years ago (in 2010). The mean LoS in this study was Comorbidity and functional impairment longer (i.e. ± 4 weeks) compared with recent clinical practice in similar SNFs. Nevertheless, we believe that these data The last aim was to focus on comorbidity and functional impairment, as both seem to play an important role in rela- reflect the current situation of geriatric stroke rehabilitation well enough, since no important changes regarding comor- tion to the occurrence of intercurrent diseases. Moreover, our results suggest that the combination of these factors bidities or intercurrent diseases are expected. Finally, comorbidity was assessed using the Charlson-CI increases the risk of developing intercurrent diseases, even more than would be expected (i.e. the ORs from the separate in relation to outcomes other than mortality, although the index was specifically designed to predict mortality. Nev - factors multiplied or summed up). This may indicate that the evaluation of comorbidity and functional status should ertheless, all detected relations showed similar results after performing sensitivity analyses in which deceased patients be integrated, preferably taking into account the functional severity of each comorbid condition. It should be noted that were excluded. some ORs were not significant due to the small size of the subgroups. A larger study is needed to further investigate this combined effect on developing intercurrent diseases dur - Conclusions ing rehabilitation. Intercurrent diseases frequently occur during geriatric stroke Strengths and limitations rehabilitation and have a detrimental effect on rehabilitation outcome, such as functional recovery and length of stay. The strength of the GRAMPS study is its multidisciplinary The present study emphasises that comorbidity and func- and multicenter prospective design in a relatively large study tional status need to be integrated and are important factors population. Whereas most studies on stroke rehabilitation associated with intercurrent diseases. In particular, diabetes investigated mainly younger patients, the present study rep- mellitus showed a strong independent association; there- resents the older, geriatric stroke patient population rela- fore, this should be a focus for screening, early detection of tively well and, therefore, strengthens external validity [27]. dysregulation and treatment, to target prevention of various The study investigated two outcomes: presence and num- intercurrent diseases. The impact of specific comorbidities ber of intercurrent diseases. Diseases were recorded using and the usefulness of routinely assessing comorbidity com- the ICD-10 coding system, and only diseases were scored bined with integrated functional severity should be further (i.e. not symptoms such as pain or dysphagia). We believe investigated. this prevents confusion regarding definitions and elucidates the role of functional activities (functional status), medical Acknowledgements The GRAMPS study was funded by two Dutch health conditions (comorbidity and intercurrent diseases) care organisations, ‘SVRZ’ in Middelburg and ‘De Zorgboog’ in Bakel (MvE and BB). The present study was supported by Laurens, a Dutch and their interactions in the complex setting of rehabilita- care organisation in Rotterdam. The authors thank Ron Wolterbeek for tion and recovery, using the ICF model as a framework [28]. his supervision regarding the statistical analyses and Laraine Visser- Another strength is the use of a Poisson regression that Isles for language editing. allowed to analyse the ‘number of intercurrent diseases’. Furthermore, we presented the classifications ‘No impact’ Compliance with ethical standards and ‘With impact’. The intercurrent diseases found in this Conflict of interest All authors declare: no external financial support study might be a selection of the more severe diseases, due from any organisation for the submitted work or any other financial to the retrospective design of registering the diseases; how- relationships with any company or organisation that might have an in- ever, analysing the impact classification as separate groups terest in the submitted work. No relationships or activities have influ- provided extra information and insight. enced the submitted work, there are no conflicts of interests. Some limitations of the study need to be considered. This Ethical approval Patients were included after the researchers received study can be considered a secondary analysis, because the a signed informed consent. GRAMPS study sample size (power) estimation was origi- nally based on the dichotomous outcome measure ‘home Informed consent Inability to give informed consent was one of the exclusion criteria. The medical ethics committee of the region discharge’, and a minimum group size of 70 was considered Nijmegen-Arnhem approved the study protocol. to be appropriate (15). However, in the present study, the groups with and without intercurrent disease were of suf- ficient size (n = 89 and n = 86, respectively). Furthermore, Open Access This article is distributed under the terms of the Crea- tive Commons Attribution 4.0 International License (http://creat iveco the cohort was a specific subgroup of older and vulnerable mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- stroke patients as presented in Supplement material Appen- tion, and reproduction in any medium, provided you give appropriate dix0020A, and data collection for the GRAMPS study ended 1 3 European Geriatric Medicine (2018) 9:347–353 353 credit to the original author(s) and the source, provide a link to the 14. Ferriero G, Franchignoni F, Benevolo E, Ottonello M, Scocchi M, Creative Commons license, and indicate if changes were made. Xanthi M (2006) The influence of comorbidities and complica- tions on discharge function in stroke rehabilitation inpatients. Eura Medicophys 42(2):91–96 15. Spruit-van EM, Buijck BI, Zuidema SU, Voncken FL, Geurts AC, Koopmans RT (2010) Geriatric rehabilitation of stroke patients in References nursing homes: a study protocol. BMC Geriatr 10:15. https://doi. org/10.1186/1471-2318-10-15 1. Palese A, Gonella S, Moreale R, Guarnier A, Barelli P, Zambiasi P 16. Koopmans RT, Lavrijsen JC, Hoek JF, Went PBM, Schols JMGA et al (2016) Hospital-acquired functional decline in older patients (2010) Dutch elderly care physician: a new generation of nursing cared for in acute medical wards and predictors: findings from a home physician specialists. J Am Geriatr Soc 58(9):1807–1809. multicentre longitudinal study. Geriatr Nurs 37(3):192–199. https https ://doi.org/10.1111/j.1532-5415.2010.03043 .x ://doi.org/10.1016/j.gerin urse0 2016.01.001 17. Denti L, Agosti M, Franceschini M (2008) Outcome predictors 2. Master Zorgmanagement. Revalidatie na een beroerte: op het of rehabilitation for first stroke in the elderly. Eur J Phys Rehabil juiste moment op de juiste plek!?: Erasmus Medical Center, http:// Med 44(1):3–11 hdl.handl e.net/2105/15643 ; 2013 [Accessed 19 Jan 2018] 18. Collin C, Wade DT, Davies S, Horne V (1988) The Barthel ADL 3. Gillick MR, Serrell NA, Gillick LS (1982) Adverse consequences index: a reliability study. Int Disabil Stud 10(2):61–63 of hospitalization in the elderly. Soc Sci Med 16(10):1033–1038 19. Sanchez-Rodriguez D, Miralles R, Muniesa JM, Mojal S, Abadia- 4. Roth EJ, Noll SF (1994) Stroke rehabilitation. 2. Comorbidities Escartin A, Vazquez-Ibar O (2015) Three measures of physical and complications. Arch Phys Med Rehabil 75(5 Spec No):S42-6 rehabilitation effectiveness in elderly patients: a prospective, lon- 5. Lew HL, Lee E, Date ES, Zeiner H (2002) Influence of medi- gitudinal, comparative analysis. BMC Geriatr 15:142. https:/ /doi. cal comorbidities and complications on FIM change and length org/10.1186/s1287 7-015-0138-5 of stay during inpatient rehabilitation. Am J Phys Med Rehabil 20. Koh GC, Chen CH, Petrella R, Thind A (2013) Rehabilitation 81(11):830–837 impact indices and their independent predictors: a systematic 6. Dromerick A, Reding M (1994) Medical and neurological review. BMJ Open 3(9):e003483. https ://doi.org/10.1136/bmjop complications during inpatient stroke rehabilitation. Stroke en-2013- 25(2):358–361 21. Charlson ME, Pompei P, Ales KL, MacKenzie CR (1987) A new 7. Roth EJ, Lovell L, Harvey RL, Heinemann AW, Semik P, Diaz method of classifying prognostic comorbidity in longitudinal stud- S (2001) Incidence of and risk factors for medical complications ies: development and validation. J Chronic Dis 40(5):373–383 during stroke rehabilitation. Stroke 32(2):523–529 22. D’Hoore W, Bouckaert A, Tilquin C (1996) Practical considera- 8. Doshi VS, Say JH, Young SH, Doraisamy P (2003) Complica- tions on the use of the Charlson comorbidity index with adminis- tions in stroke patients: a study carried out at the rehabilita- trative data bases. J Clin Epidemiol 49(12):1429–1433 tion medicine service, Changi General Hospital. Singap Med J 23. Goldstein LB, Samsa GP, Matchar DB, Horner RD (2004) Charl- 44(12):643–652 son Index comorbidity adjustment for ischemic stroke outcome 9. Hung JW, Tsay TH, Chang HW, Leong CP, Lau YC (2005) Inci- studies. Stroke Journal of Cerebral Circulation 35(8):1941–1945 dence and risk factors of medical complications during inpatient 24. Knol MJ, Groenwold RH (2015) Effect modification and interac- stroke rehabilitation. Chang Gung Med J 28(1):31–38 tion. Ned Tijdschr Geneeskd 159:A8499 10. Kuptniratsaikul V, Kovindha A, Suethanapornkul S, Manim- 25. Spruit-van EM, Zuidema SU, Buijck BI, Koopmans RT, Geurts manakorn N, Archongka Y (2009) Complications during the AC (2012) Determinants of rehabilitation outcome in geriatric rehabilitation period in Thai patients with stroke: a multicenter patients admitted to skilled nursing facilities after stroke: a Dutch prospective study. Am J Phys Med Rehabil 88(2):92–99. https :// multi-centre cohort study. Age Ageing 41(6):746–752 doi.org/10.1097/PHM.0b013 e3181 909d5 f 26. Gokkaya N, Aras M, Cardenas D, Kaya A (2006) Stroke reha- 11. Kalra L, Yu G, Wilson K, Roots P (1995) Medical complications bilitation outcome: the Turkish experience. Int J Rehabil Res during stroke rehabilitation. Stroke 26(6):990–994 29(2):105–111 12. Civelek GM, Atalay A, Turhan N (2016) Medical complications 27. Gaynor EJ, Geoghegan SE, O’Neill D (2014) Ageism in stroke experienced by first-time ischemic stroke patients during inpatient, rehabilitation studies. Age Ageing 43(3):429–431. https ://doi. tertiary level stroke rehabilitation. J Phys Ther Sci 28(2):382–391. org/10.1093/agein g/afu02 6 https ://doi.org/10.1589/jpts.28.382 Epub 2016 Feb 29 28. Gladman JR (2008) The international classification of functioning, 13. Siegler EL, Stineman MG, Maislin G (1994) Development disability and health and its value to rehabilitation and geriatric of complications during rehabilitation. Arch Intern Med medicine. J Chin Med Assoc JCMA 71(6):275–278 154(19):2185–2190 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Geriatric Medicine Springer Journals

Comorbidity and intercurrent diseases in geriatric stroke rehabilitation: a multicentre observational study in skilled nursing facilities

Free
7 pages

Loading next page...
 
/lp/springer_journal/comorbidity-and-intercurrent-diseases-in-geriatric-stroke-c8AU1fyeVN
Publisher
Springer International Publishing
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Geriatrics/Gerontology; Internal Medicine
eISSN
1878-7657
D.O.I.
10.1007/s41999-018-0043-5
Publisher site
See Article on Publisher Site

Abstract

Background Older patients often have multiple comorbidities and are susceptible to develop intercurrent diseases during rehabilitation. This study investigates intercurrent diseases and associated factors in patients undergoing geriatric stroke rehabilitation, focussing on pre-existing comorbid conditions, overall comorbidity and baseline functional status. Materials and methods This multicentre prospective cohort study included 15 skilled nursing facilities. Data were collected at baseline and at discharge. The primary outcome measures were presence and number of intercurrent diseases. Further- more, their impact on change in rehabilitation goals or length of stay was examined. Comorbidity was assessed with the Charlson index, and functional status with the Barthel index (BI). Results Of the 175 included patients, 51% developed an intercurrent disease. A lower baseline BI, a higher Charlson index, presence of diabetes mellitus (DM) and kidney disease were related to the occurrence of an intercurrent disease (p < 0.05). Moreover, a lower BI, a higher Charlson index, and particularly the presence of DM were independently associated. If both comorbidity and a lower baseline functional status were present, the odds ratio (95% CI) of developing intercurrent diseases was 6.70 [2.33–19.2], compared to 1.73 [0.52–5.72] (comorbidity only) and 1.62 [0.53–4.94] (only BI ≤ 14). Conclusions On admission, functional impairments and comorbidity, particularly diabetes, independently contribute to developing intercurrent diseases during geriatric stroke rehabilitation. Therefore, routine evaluation of comorbidity integrated with functional status at the start of rehabilitation is essential to identify patients at risk. Finally, particular attention should be paid to patients with DM to prevent intercurrent diseases and support optimal functional recovery . Keywords Geriatric rehabilitation · Comorbidity · Functional impairments · Intercurrent diseases · Stroke Introduction Electronic supplementary material The online version of this article (https ://doi.org/10.1007/s4199 9-018-0043-5) contains supplementary material, which is available to authorized users. Following acute hospitalisation, rehabilitation helps patients to regain functional independency that enables them to be * Anouk D. Kabboord discharged home. However, during hospitalisation, the A.D.Kabboord@lumc.nl risk of functional decline and complications is particularly Department of Public Health and Primary Care, Leiden increased in older patients [1]. In the Netherlands, about University Medical Center, Hippocratespad 21, Postbus one-third of all stroke patients are referred to a skilled nurs- 9600, 2300 RC Leiden, The Netherlands ing facility (SNF) that provides geriatric rehabilitation. Department of Neurology, Erasmus MC University Medical These patients are usually relatively older, have a longer Center, ‘s-Gravendijkwal 230, Rotterdam, The Netherlands length of stay (LoS) in the acute hospital, and have more Department of Primary and Community Care, Centre complex problems (Supplement material Appendix A) [2]. for Family Medicine, Geriatric Care and Public Health, Also, during inpatient rehabilitation, intercurrent diseases Radboud University Nijmegen Medical Centre, Nijmegen, may occur that interfere with therapy and could negatively The Netherlands 4 impact rehabilitation outcome [3, 4]. “Joachim en Anna”, Centre for Specialized Geriatric Care, Nijmegen, The Netherlands Vol.:(0123456789) 1 3 348 European Geriatric Medicine (2018) 9:347–353 Studies investigating complications during inpatient prolonged the LoS or (ii) whether the rehabilitation goals stroke rehabilitation found that 30–96% of the patients needed adjustment. Four categories were formed: (1) no developed complications; this wide range could be due to intercurrent disease, (2) ‘No impact’, (3) ‘With impact’, and different definitions of a complication and the methods of (4) intercurrent disease that directly caused death. measurement [5–12]. The present study investigates inter- current diseases, i.e. any disease that occurs during the Data collection progress of another disease, during rehabilitation. Factors related to intercurrent diseases can include age, gender [9, The participating multidisciplinary teams consisted of a 13], time interval between stroke and rehabilitation [7, 10, physician [16], a physiotherapist, an occupational therapist, 12], severe stroke [7, 11] or functional impairment [6, 9, 12, a psychologist, a speech therapist, a dietician and skilled 13] and comorbidity [5–7, 13, 14], although it is unknown nurses; all received the same instructions regarding perfor- which specific comorbidities are related. Particularly, older mance of the assessments. Data were collected within the patients are at risk of functional decline and often have mul- first 2 weeks after admission (T0) and at discharge (T1) from tiple comorbidities. However, few studies have investigated the SNF or (at the latest) 1 year after admission, if a patient associations with intercurrent diseases in the older, vulner- was still in the SNF at that time. able group of patients receiving geriatric stroke rehabilita- tion [5, 13]. Furthermore, intercurrent diseases may impede Measurements successful functional recovery [5, 14]. Therefore, to better understand the relations between comorbidity, functional The following patient characteristics and data were col- impairment and intercurrent diseases, and to identify asso- lected: age, gender, home situation, comorbidity, LoS in ciated pre-existing comorbid conditions, this study explores: acute hospital, LoS in the SNF, and discharge destination (i) the presence, and number of intercurrent diseases and [5–14, 17]. Functional assessment was performed at base- their impact on older patients admitted to an SNF, recover- line and at discharge using the modified Barthel index (BI) ing after stroke, and (ii) factors associated with the presence to assess activities of daily living (ADL) [18]. Premorbid and number of intercurrent diseases, focusing on functional BI was assessed on admission, using information on the status and comorbidity. patient’s situation prior to the acute stroke, based on inter- view and collateral history. Functional recovery was defined in two ways: BI at discharge and ‘relative functional gain’, Materials and methods which was calculated as follows: (BI-discharge minus BI- admission)/(BI-premorbid minus BI-admission) × 100 [19, Participants 20]. Relative functional gain expresses the achieved percent- age of potential functional gain. Data were obtained from the Geriatric Rehabilitation in Pre-existing comorbidity was assessed using the Charl- AMPutation and Stroke (GRAMPS) study. Data collection son comorbidity index (Charlson-CI). This index consists took place between January 2008 and July 2010; details of 19 diagnoses and was adjusted for stroke [21–23]. The on the study design are already published [15]. A total of Charlson-CI was categorised as: 0 (no comorbidity), 1 (sin- 15 SNFs located in the southern part of the Netherlands gle comorbidity) or ≥ 2 (multiple comorbidities), unless oth- participated. All stroke patients admitted to one of these erwise specified. Comorbidities were recorded if present in SNFs were eligible for inclusion. Patients were excluded medical history, e.g. chronic diseases and conditions that if they refused participation, were unable to give informed required ongoing use of (preventive) medication. Conditions consent, were critically ill, or were expected to have a stay that had completely resolved without any residual symp- of ≤ 2 weeks. The medical ethics committee of the region toms or need for treatment were not noted (e.g. childhood Nijmegen-Arnhem approved the study protocol. asthma). Finally, if myocardial infarction in the past had led to heart failure, only heart failure was recorded. Outcome measures Statistical analysis For the present study, the outcome measures were: the pres- ence and number of intercurrent diseases that occurred dur- Data were processed and analysed using the Statistical Pack- ing rehabilitation. Intercurrent diseases were coded using the 10th revision Clinical Modification ICD-10CM. At dis- age for Social Science version 23. Means with standard deviations (normal distribution), medians with interquartile charge, the attending physician registered intercurrent dis- eases that affected the course of the rehabilitation: impact ranges (skewed data), or absolute numbers with percentages (categorical data) are reported. was classified according to (i) whether the disease had 1 3 European Geriatric Medicine (2018) 9:347–353 349 A Chi-Squared test (categorical data), ANOVA or Characteristics related to intercurrent diseases Kruskal–Wallis test, depending on their distribution, were used to detect mean differences in characteristics Patients without any intercurrent disease had a BI on admis- between the four intercurrent disease categories and to sion of at least 4 points higher than those with intercurrent identify comorbid conditions related to the occurrence of diseases. The proportion of patients without comorbidity was intercurrent diseases. A p value of ≤ 0.05 was considered largest in the category ‘no intercurrent disease’ (52%), whereas statistically significant. in the category ‘With impact’, the proportion of patients with Multivariate analyses were performed using binary multiple comorbidities was the largest (54%), p = 0.007. logistic regression with the presence of intercurrent dis- Patients that developed intercurrent diseases were less often eases and Poisson regression with number of intercur- discharged home, had a longer LoS, a lower BI at discharge, rent diseases as the dependent variable. Rehabilitation and a lower relative functional gain. This also applied to the LoS (log) was added as the ‘offset’. Factors included in category that was considered as having ‘No impact’. Mul- the multivariate model were age and gender. Significant tivariate analyses showed that: BI on admission (OR 0.87 baseline variables (p < 0.10) were added as a continuous [0.82–0.92]) and comorbidity (OR 1.43 [1.13–1.81]) were variable if applicable. independently associated with the presence of intercurrent Before performing the analyses, data were tested for diseases, but only the Charlson-CI was significantly associated the required assumptions, such as multicollinearity, inter- with number of intercurrent diseases (incidence rate ratio: 1.14 action and effect modification. To investigate comorbid- [1.03–1.25], p: 0.008). This means that with every extra point ity and baseline functional status, separate and combined on the Charlson-CI, a 14% increase in the number of intercur- relations with the presence of intercurrent diseases were rent diseases is expected (Supplement material Appendix D). analysed. For this purpose, variables were dichotomized. Odds ratios (OR) were calculated with the absence of both Comorbidity and intercurrent diseases factors as reference category [24]. Sensitivity analyses were performed, i.e. with and without deceased patients. Having diabetes and/or kidney disease was significantly related to the occurrence of an intercurrent disease (Table 2). More- over, logistic regression analysis showed that only diabetes was independently associated (OR 3.50 [1.32–9.26]). No clear Results patterns or relations between comorbidities and specific inter - current diseases were observed: a wide variety of different Characteristics diseases occurred in patients with pre-existing comorbidity. The intercurrent diseases that most frequently occurred were Of the 378 eligible patients, 186 were included in the cardiovascular (13%), psychiatric (12%) such as depression GRAMPS study; the excluded patients did not differ and delirium, and genitourinary (11%), predominantly uri- with regard to age, gender or LoS [25]. The present study nary tract infections. An overview of intercurrent diseases, included 175 patients because 11 patients were lost to per comorbidity (the five most prevalent only), is presented in follow-up, mainly due to translocation to another SNF Supplement material Appendix E. (Supplement material Appendices B and C). Table 1 pre- sents the baseline characteristics of the study population, Comorbidity and baseline functional status and the intercurrent disease categories. Mean age was 78.8 years and 46% were males. On average, LoS in the Table 3 shows the cumulative effect of the combination of a acute hospital was 19  days, the premorbid BI was 20, lower functional status on admission (BI ≤ 14) and the pres- baseline BI was 12, and BI at discharge was 17. LoS in ence of comorbidity (Charlson-CI ≥ 1) in relation to the occur- the SNF was 12 weeks, the (average) relative functional rence of an intercurrent disease. On admission, when comor- gain was 67, and 56% of these patients was discharged bidity and lower functional status on admission were present home. separately, ORs were 1.73 [0.52–5.72] and 1.62 [0.53–4.94], Of the 89 (51%) patients that developed an intercurrent respectively. However, if both were present, the OR was 6.70 disease, 49% developed one disease, 33% ≥ 2 diseases, [2.33–19.2]. and 18% died. Comorbidity was present in 116 (62%) patients: 40 (21%) scored 1 and 76 (41%) scored ≥ 2. The most prevalent pre-existing comorbidities were myocar- dial infarction (18%), diabetes mellitus (18%) and conges- tive heart failure (16%). 1 3 350 European Geriatric Medicine (2018) 9:347–353 Table 1 Patient characteristics classified by intercurrent disease (ID) impact category Total baseline ID absent ID no impact ID with impact ID deceased ID impact n = 175 n = 86 n = 22 n = 46 n = 16 unknown n = 5 Variables at baseline Age (years), Mean 78.8 (8.0) 78.2 (8.3) 78.8 (5.6) 78.9 (8.5) 81.2 (8.4) 82.6 (7.8) (SD) Gender (male), n (%) 80 (46) 45 (52) 11 (50) 16 (35) 7 (44) 1 (20) Charlson-CI score, 1 (2)* 0 (2) 1 (2) 2 (2) 2 (2) 2 (3) median (IQR) Charlson-CI = 0, n 68 (39) 45 (52) 8 (36) 10 (22) 3 (19) 2 (40) (%) Charlson-CI = 1, n 38 (22) 19 (22) 6 (27) 11 (24) 2 (13) 0 (0) (%) Charlson-CI ≥ 2, n 69 (39) 22 (26) 8 (36) 25 (54) 11 (69) 3 (60) (%) Premorbid Barthel 20 (3) 20 (2) 20 (2) 19 (3) 17 (7) 18 (3) Index, median (IQR) LoS acute hospital in 19 (14) 19 (11) 19 (13) 19.5 (18) 22 (18) 21 (21) days, median (IQR) Barthel Index on 12 (10)* 14 (7) 9 (12) 9 (8) 8 (9) 10 (6) admission, median (IQR) Variables at discharge LoS rehabilitation 12 (15)* 8 (6) 16 (23) 22 (26) – 16 (6) in weeks, median (IQR) Barthel Index at 17 (8)* 18 (4) 16 (9) 11 (10) – 15 (4) discharge, median (IQR) Relative functional 67 (90)* 85 (84) 67 (76) 24 (79) – 71 (42) gain, median (IQR) Discharge home, n 88 (56) 62 (73) 9 (43) 13 (28) – 4 (80) (%) Equal statistical significance was found when deceased patients were excluded SD standard deviation, Charlson-CI Charlson comorbidity index, IQR interquartile range, LoS length of stay Statistical significance at p < 0.05 *Kruskal–Wallis test Chi-Square test functional status, higher pre-existing comorbidity burden in Discussion general and specifically the presence of diabetes mellitus were independent determinants of developing intercurrent Main findings diseases. Furthermore, patients with multiple comorbidi- ties (higher Charlson-CI) had an increased risk to develop a To our knowledge, this is the first study to focus on comor - higher number of intercurrent diseases. Finally, the odds of bidity and intercurrent diseases during geriatric stroke reha- developing an intercurrent disease were substantially higher bilitation. The study cohort was characterised by a large drop if a patient had both comorbidity and functional impairment in functional status after acute stroke, often with multiple than if only one of these factors were present. comorbidities and a higher age compared to the majority of studies on stroke patients [5–10, 12–14, 26]. Although this subgroup had been triaged for inpatient geriatric reha- Intercurrent diseases bilitation, and selected as a vulnerable subgroup of patients on the base of medical complexity and functional depend- The percentage of patients (51%) that developed inter- ency, discriminant factors were still present. Lower baseline current diseases is comparable to that of studies using an 1 3 European Geriatric Medicine (2018) 9:347–353 351 Table 2 Associations between Comorbidity Total ID absent ID present comorbid conditions and # # presence of ≥ 1 intercurrent Charlson index, median (IQR) 1 (2) 0 (2) 2 (3) disease (ID) Comorbid condition Myocardial infarction, n (%) 31 (18) 13 18 Heart failure, n (%) 29 (17) 10* 19* Peripheral vascular disease, n (%) 23 (13) 9 14 Dementia, n (%) 1 (1) 1 0 Chronic pulmonary disease, n (%) 18 (10) 8 10 Musculoskeletal/connective tissue, n (%) 9 (5) 2* 7* Ulcers, n (%) 8 (5) 2 6 Mild liver disease, n (%) 3 (2) 1 2 # # Kidney disease (moderate), n (%) 16 (9) 3 13 # # Diabetes mellitus, n (%) 31 (18) 9 22 Malignancy, n (%) 10 (6) 3 7 Leukaemia, n (%) 1 (1) 1 0 Lymphoma, n (%) 2 (1) 0 2 Moderate liver disease, n (%) 0 (0) 0 0 Metastasis of solid tumour, n (%) 3 (2) 1 2 Any malignancy (of the above mentioned), n (%) 13 (7) 4 9 Comorbidities included in the logistic regression analysis are presented in bold Chi Square test: * p < 0.10, p < 0.05 Table 3 Comorbidity and baseline function: the separate and com- rates were similar to those in studies using prospective bined effect on developing an intercurrent disease in geriatric stroke assessment and similar prevalent diseases were found, i.e. rehabilitation (n = 170) genitourinary (urinary tract infections) and psychiatric dis- Charlson-CI BI ≤ 14 on Intercurrent Odds ratio [95% CI] eases (depression and delirium) [6–12, 14, 17]. However, score ≥ 1 admission* disease in the present study intercurrent cardiovascular disease was more prevalent, presumably because pre-existing Yes No cardiovascular comorbidities were highly prevalent in our No No 6 17 Reference 1.00 subgroup of vulnerable geriatric patients. No Yes 16 28 1.62 [0.53–4.94] Yes No 11 18 1.73 [0.52–5.72] Yes Yes 52 22 6.70 [2.33–19.2] Intercurrent diseases and their associations Charlson-CI charlson comorbidity index, BI Barthel index; CI confi- dence interval The presence of intercurrent diseases was related to reha- *Assessing the BI on admission was not possible in 5 patients. Sen- bilitation impact indices (longer LoS, less functional sitivity analysis showed similar results: when deceased patients were recovery and less often being discharged home). Despite excluded (n = 154) ORs were 1.32 [0.42–4.11], 1.42 [0.42–4.83] and that physicians retrospectively registered intercurrent dis- 5.54 [1.91–16.0], respectively eases according to their influence on rehabilitation, it was striking that this relation also applied to the category ‘No assessment method similar to ours (i.e. 30–54%) [5, 8, impact’. This underlines the impact that intercurrent dis- eases can have on rehabilitation outcomes. Besides base- 9, 13, 14, 17]. However, although other studies found a higher rate (60–100%), there was a clear difference in the line functional status and comorbidity in general, diabetes mellitus was found to be a significant determinant of the methods used. For example, shoulder pain, limb spasticity, dysphagia or aphasia were categorised as a complication, occurrence of an intercurrent disease. Diabetes affects various organ systems (e.g. vascular, skin, eyes, nervous whereas in the present study (and similar studies) these were considered to be symptoms and not diseases [6, 7, system) and might be the (underlying) cause of a variety of intercurrent diseases. However, the present study had 10–12, 26]. In this study, we were specifically interested in intercurrent diseases that occurred during the inpatient insufficient power to further investigate different comor - bidities and their associations with specific intercurrent rehabilitation period, and physicians retrospectively regis- tered the intercurrent diseases. Nevertheless, our incidence diseases. 1 3 352 European Geriatric Medicine (2018) 9:347–353 some years ago (in 2010). The mean LoS in this study was Comorbidity and functional impairment longer (i.e. ± 4 weeks) compared with recent clinical practice in similar SNFs. Nevertheless, we believe that these data The last aim was to focus on comorbidity and functional impairment, as both seem to play an important role in rela- reflect the current situation of geriatric stroke rehabilitation well enough, since no important changes regarding comor- tion to the occurrence of intercurrent diseases. Moreover, our results suggest that the combination of these factors bidities or intercurrent diseases are expected. Finally, comorbidity was assessed using the Charlson-CI increases the risk of developing intercurrent diseases, even more than would be expected (i.e. the ORs from the separate in relation to outcomes other than mortality, although the index was specifically designed to predict mortality. Nev - factors multiplied or summed up). This may indicate that the evaluation of comorbidity and functional status should ertheless, all detected relations showed similar results after performing sensitivity analyses in which deceased patients be integrated, preferably taking into account the functional severity of each comorbid condition. It should be noted that were excluded. some ORs were not significant due to the small size of the subgroups. A larger study is needed to further investigate this combined effect on developing intercurrent diseases dur - Conclusions ing rehabilitation. Intercurrent diseases frequently occur during geriatric stroke Strengths and limitations rehabilitation and have a detrimental effect on rehabilitation outcome, such as functional recovery and length of stay. The strength of the GRAMPS study is its multidisciplinary The present study emphasises that comorbidity and func- and multicenter prospective design in a relatively large study tional status need to be integrated and are important factors population. Whereas most studies on stroke rehabilitation associated with intercurrent diseases. In particular, diabetes investigated mainly younger patients, the present study rep- mellitus showed a strong independent association; there- resents the older, geriatric stroke patient population rela- fore, this should be a focus for screening, early detection of tively well and, therefore, strengthens external validity [27]. dysregulation and treatment, to target prevention of various The study investigated two outcomes: presence and num- intercurrent diseases. The impact of specific comorbidities ber of intercurrent diseases. Diseases were recorded using and the usefulness of routinely assessing comorbidity com- the ICD-10 coding system, and only diseases were scored bined with integrated functional severity should be further (i.e. not symptoms such as pain or dysphagia). We believe investigated. this prevents confusion regarding definitions and elucidates the role of functional activities (functional status), medical Acknowledgements The GRAMPS study was funded by two Dutch health conditions (comorbidity and intercurrent diseases) care organisations, ‘SVRZ’ in Middelburg and ‘De Zorgboog’ in Bakel (MvE and BB). The present study was supported by Laurens, a Dutch and their interactions in the complex setting of rehabilita- care organisation in Rotterdam. The authors thank Ron Wolterbeek for tion and recovery, using the ICF model as a framework [28]. his supervision regarding the statistical analyses and Laraine Visser- Another strength is the use of a Poisson regression that Isles for language editing. allowed to analyse the ‘number of intercurrent diseases’. Furthermore, we presented the classifications ‘No impact’ Compliance with ethical standards and ‘With impact’. The intercurrent diseases found in this Conflict of interest All authors declare: no external financial support study might be a selection of the more severe diseases, due from any organisation for the submitted work or any other financial to the retrospective design of registering the diseases; how- relationships with any company or organisation that might have an in- ever, analysing the impact classification as separate groups terest in the submitted work. No relationships or activities have influ- provided extra information and insight. enced the submitted work, there are no conflicts of interests. Some limitations of the study need to be considered. This Ethical approval Patients were included after the researchers received study can be considered a secondary analysis, because the a signed informed consent. GRAMPS study sample size (power) estimation was origi- nally based on the dichotomous outcome measure ‘home Informed consent Inability to give informed consent was one of the exclusion criteria. The medical ethics committee of the region discharge’, and a minimum group size of 70 was considered Nijmegen-Arnhem approved the study protocol. to be appropriate (15). However, in the present study, the groups with and without intercurrent disease were of suf- ficient size (n = 89 and n = 86, respectively). Furthermore, Open Access This article is distributed under the terms of the Crea- tive Commons Attribution 4.0 International License (http://creat iveco the cohort was a specific subgroup of older and vulnerable mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- stroke patients as presented in Supplement material Appen- tion, and reproduction in any medium, provided you give appropriate dix0020A, and data collection for the GRAMPS study ended 1 3 European Geriatric Medicine (2018) 9:347–353 353 credit to the original author(s) and the source, provide a link to the 14. Ferriero G, Franchignoni F, Benevolo E, Ottonello M, Scocchi M, Creative Commons license, and indicate if changes were made. Xanthi M (2006) The influence of comorbidities and complica- tions on discharge function in stroke rehabilitation inpatients. Eura Medicophys 42(2):91–96 15. Spruit-van EM, Buijck BI, Zuidema SU, Voncken FL, Geurts AC, Koopmans RT (2010) Geriatric rehabilitation of stroke patients in References nursing homes: a study protocol. BMC Geriatr 10:15. https://doi. org/10.1186/1471-2318-10-15 1. Palese A, Gonella S, Moreale R, Guarnier A, Barelli P, Zambiasi P 16. Koopmans RT, Lavrijsen JC, Hoek JF, Went PBM, Schols JMGA et al (2016) Hospital-acquired functional decline in older patients (2010) Dutch elderly care physician: a new generation of nursing cared for in acute medical wards and predictors: findings from a home physician specialists. J Am Geriatr Soc 58(9):1807–1809. multicentre longitudinal study. Geriatr Nurs 37(3):192–199. https https ://doi.org/10.1111/j.1532-5415.2010.03043 .x ://doi.org/10.1016/j.gerin urse0 2016.01.001 17. Denti L, Agosti M, Franceschini M (2008) Outcome predictors 2. Master Zorgmanagement. Revalidatie na een beroerte: op het of rehabilitation for first stroke in the elderly. Eur J Phys Rehabil juiste moment op de juiste plek!?: Erasmus Medical Center, http:// Med 44(1):3–11 hdl.handl e.net/2105/15643 ; 2013 [Accessed 19 Jan 2018] 18. Collin C, Wade DT, Davies S, Horne V (1988) The Barthel ADL 3. Gillick MR, Serrell NA, Gillick LS (1982) Adverse consequences index: a reliability study. Int Disabil Stud 10(2):61–63 of hospitalization in the elderly. Soc Sci Med 16(10):1033–1038 19. Sanchez-Rodriguez D, Miralles R, Muniesa JM, Mojal S, Abadia- 4. Roth EJ, Noll SF (1994) Stroke rehabilitation. 2. Comorbidities Escartin A, Vazquez-Ibar O (2015) Three measures of physical and complications. Arch Phys Med Rehabil 75(5 Spec No):S42-6 rehabilitation effectiveness in elderly patients: a prospective, lon- 5. Lew HL, Lee E, Date ES, Zeiner H (2002) Influence of medi- gitudinal, comparative analysis. BMC Geriatr 15:142. https:/ /doi. cal comorbidities and complications on FIM change and length org/10.1186/s1287 7-015-0138-5 of stay during inpatient rehabilitation. Am J Phys Med Rehabil 20. Koh GC, Chen CH, Petrella R, Thind A (2013) Rehabilitation 81(11):830–837 impact indices and their independent predictors: a systematic 6. Dromerick A, Reding M (1994) Medical and neurological review. BMJ Open 3(9):e003483. https ://doi.org/10.1136/bmjop complications during inpatient stroke rehabilitation. Stroke en-2013- 25(2):358–361 21. Charlson ME, Pompei P, Ales KL, MacKenzie CR (1987) A new 7. Roth EJ, Lovell L, Harvey RL, Heinemann AW, Semik P, Diaz method of classifying prognostic comorbidity in longitudinal stud- S (2001) Incidence of and risk factors for medical complications ies: development and validation. J Chronic Dis 40(5):373–383 during stroke rehabilitation. Stroke 32(2):523–529 22. D’Hoore W, Bouckaert A, Tilquin C (1996) Practical considera- 8. Doshi VS, Say JH, Young SH, Doraisamy P (2003) Complica- tions on the use of the Charlson comorbidity index with adminis- tions in stroke patients: a study carried out at the rehabilita- trative data bases. J Clin Epidemiol 49(12):1429–1433 tion medicine service, Changi General Hospital. Singap Med J 23. Goldstein LB, Samsa GP, Matchar DB, Horner RD (2004) Charl- 44(12):643–652 son Index comorbidity adjustment for ischemic stroke outcome 9. Hung JW, Tsay TH, Chang HW, Leong CP, Lau YC (2005) Inci- studies. Stroke Journal of Cerebral Circulation 35(8):1941–1945 dence and risk factors of medical complications during inpatient 24. Knol MJ, Groenwold RH (2015) Effect modification and interac- stroke rehabilitation. Chang Gung Med J 28(1):31–38 tion. Ned Tijdschr Geneeskd 159:A8499 10. Kuptniratsaikul V, Kovindha A, Suethanapornkul S, Manim- 25. Spruit-van EM, Zuidema SU, Buijck BI, Koopmans RT, Geurts manakorn N, Archongka Y (2009) Complications during the AC (2012) Determinants of rehabilitation outcome in geriatric rehabilitation period in Thai patients with stroke: a multicenter patients admitted to skilled nursing facilities after stroke: a Dutch prospective study. Am J Phys Med Rehabil 88(2):92–99. https :// multi-centre cohort study. Age Ageing 41(6):746–752 doi.org/10.1097/PHM.0b013 e3181 909d5 f 26. Gokkaya N, Aras M, Cardenas D, Kaya A (2006) Stroke reha- 11. Kalra L, Yu G, Wilson K, Roots P (1995) Medical complications bilitation outcome: the Turkish experience. Int J Rehabil Res during stroke rehabilitation. Stroke 26(6):990–994 29(2):105–111 12. Civelek GM, Atalay A, Turhan N (2016) Medical complications 27. Gaynor EJ, Geoghegan SE, O’Neill D (2014) Ageism in stroke experienced by first-time ischemic stroke patients during inpatient, rehabilitation studies. Age Ageing 43(3):429–431. https ://doi. tertiary level stroke rehabilitation. J Phys Ther Sci 28(2):382–391. org/10.1093/agein g/afu02 6 https ://doi.org/10.1589/jpts.28.382 Epub 2016 Feb 29 28. Gladman JR (2008) The international classification of functioning, 13. Siegler EL, Stineman MG, Maislin G (1994) Development disability and health and its value to rehabilitation and geriatric of complications during rehabilitation. Arch Intern Med medicine. J Chin Med Assoc JCMA 71(6):275–278 154(19):2185–2190 1 3

Journal

European Geriatric MedicineSpringer Journals

Published: Mar 13, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off