Many older adults do not return to their premorbid health states following hospital admissions requiring a period of subacute inpatient rehabilitation. Yet, the efficacy of inpatient rehabilitation specifically designed for geriatric patients is supported by the literature   . Valid results can be achieved notwithstanding the necessity to overcome adverse or unexpected clinical events, and to manage conditions like frailty and comorbidity      . Frail older patients face complex, multidimensional problems that specialized geriatric services have to address, in a comprehensive, multidisciplinary approach.
Outcome selection in patient care includes many goals, not all of which will be relevant for all patients  . Hence it may be beneficial for patients to have an accurate understanding of their likely recovery in order to make informed decisions about shared therapy goals and when planning for living in the community after discharge from a rehabilitative setting. Accurate understanding of likely therapeutic benefit and the personal capacity for improvement is constructive for informing joint decision making between patients and health professionals and to facilitate adherence to therapeutic protocols. Such vision stands at the core of person-centered models of care that are becoming increasingly common among subacute rehabilitation facilities. Such models of care may assist older patients recovering from illness or injury to work with their treating team to focus therapy and other treatments on deficits that are most important for them, in order to optimize their health-related quality of life at discharge or thereafter  .
Yet, literature on matching between patients’ expectations and their rehabilitative outcomes is scanty.
2. Paper’s Purpose
The aim of our study was twofold: 1) to investigate clients’ expectations about outcomes from staying in our geriatric rehabilitation ward, and 2) to check the matching between such expectation and actual outcomes.
3. Material and Methods
The study is designed as retrospective evaluation of rehabilitative outcome of elderly patients, matched to their respective expectations.
Survey on ordinary activity within a rehabilitative/intermediate care ward. Due to the exploratory and retrospective nature of our enquiry, neither sample size calculation nor selection (inclusion/exclusion) criteria were applied. Data pertain to patients consecutively admitted to the ward since September 2012 to march 2014.
Data used for analysis were collected as part of routine clinical practice: therefore formal approval by the local ethics committee was not sought.
3.2. Outcome and Explanatory Measures
Expectations were drawn from an open field within an on purpose, closed questions questionnaire, delivered to clients at entry. The self-made questionnaire taps issues relevant to the stay [e.g.: relationships with staff, privacy, food]; it is delivered at entry and recollected after about one week, in order to give patients time to get acclimatized to the ward. Free lines are available in the format to answer the explicit question: “What do You expect from your rehabilitative stay?”.
Expectations answers were coded in a two step process: the first step yielded 9 main categories, that were subsequently brought into 5 broad categories: functional improvement, clinical improvement, overall improvement, discharge to home, no expectation [either implicitly or explicitly].
Functional, clinical and outcome information were registered through a standardized data collection form. Such data collection system has been previously validated elsewhere  . After general and demographic variables, it comprehends typical measures of multidimensional assessment, including the pre-morbid modified Rankin Scale (mRS)  , the Barthel Index (BI)  , and the Mini-Mental State Examination (MMSE)   . Besides the BI, also performances through POMA test by Tinetti et al. were measured both at entry and at discharge   . A core set of dichotomous standardized indicators and validated assessment tools follows in the collection form. Specifically, the conditions with nursing needs―assessed at admission and at discharge in a dichotomous checklist― were: dysphagia, malnutrition, nasogastric tube or percutaneus endoscopic gastrostomy, pressure sore, indwelling bladder catheter, urinary incontinence, central venous catheter and tracheostomy  .
Sociodemographic, functional and clinical characteristics of the participants are summarized in Table 1. Briefly, 186 patients consecutively were admitted to the ward since fall 2012 to spring 2014: some adult people are included in a sample predominantly constituted by elderly persons. Primary reasons for admission spanned through a wide range of diseases: muscolar-skeketal disorders [either fractures―mostly femur fractures (18%)―or degenerative diseases, including elective arthroplasty (21% overall)], neurologic impairments [stroke (11%), Parkinson’s disease and parkinsonism (3%), neuropathies (5%)], deconditioning syndromes due to various medical and/or surgical conditions (25%). Overall, the
Table 1. Sociodemographic, functional and clinical characteristics of the participants.
*EPUAP: European Pressure Ulcer Advisory Panel classification  .
sample was characterized―as an administrative admission requisite―by pre-mor- bid physical and or cognitive frailty or partial dependency, and/or co-morbidity, as shown by pre-morbid Barthel Index total score. At entry, 3 out of 5 patients showed poor cognitive performance, according to MMSE total score classes. Clinical conditions were often compromised at admission: beyond what reported into Table 1, 8% of patients active tumor; 7% had dysphagia, 13 hosted an% indwelling urinary catheter; 15% were clinically instable according to Hulm criteria  ; 9% had moderate to severe dementia and 5% delirium―assessed according to Confusion Assessment Method―  ; 48% were clinically depressed.
3.4. Statistical Analysis
Descriptive and inferential analysis are performed according to normal or
not-normal data distributions; the SPSS software was employed.
Recoded expectations were distributed as follows: functional improvement 31.4%, overall improvement 29.6%, clinical improvement 21.5%, home discharge 7%; 7.5% of answers were blank or explicitly declared having no expectations.
No baseline characteristics differed according to expectations, except for MMSE values, which were lower for patients whose answers were blank or explicitly declared having no expectations [14.9/30 (s.d. 8,3/30) and higher for people aiming primarily to go back home [25.1/30 (s.d. 3.9/30)] (p = 0.013).
Neither length of stay nor rehabilitative sessions numbers differed according to expectations.
As shown in Table 2, the case mix at entry is featured by high levels of dependency in basic activities of daily life, with special regards to walking ability; further, patients at admission were usually charged by one condition with nursing needs. All these three variables significantly improved at discharge; overall dependency in basic activities of daily living shifted by one category according to total Barthel Index score. At discharge, most people were able to walk independently in short range (50 meters), with supervision and/or walking aid if needed [see median and modal value of locomotion item at Barthel Index]. Most people were discharged free of conditions requiring nursing needs.
4.3. Expectations/Outcomes Matching
In one case the match between expectations and actual outcomes was fulfilled. Improvements in conditions with nursing needs was highest in patients whose expectations were clinical improvement [p = 0.029 (Brown-Forsythe)]. Further, highest improvements in walking ability (6.2 points at specific Barthel Index item) were achieved by patients prioritizing a clinical improvement, followed by those aiming at functional as well as overall improvement (5 points at specific Barthel Index item) [p = 0.009 (Welch); p = 0.041 (Brown-Forsythe)]. In both cases, worst outcomes were got by patients without implicit or explicit expectations: smallest improvements in walking ability (2.8 points at specific Barthel Index item) or stability in conditions requiring nursing needs, respectively. No further statistically significant matching were found between expectations and outcomes. The number of rehabilitative sessions attended did not vary according to expectations.
In previous works, we got valuable rehabilitative results, even in case of cognitive impairment, pursuing our geriatric paradigm [to “select” (true from irrelevant) and to “hierarchy” (top/bottom) problems, in order to link together interventions and solutions]  . Here we focused onto patient’s expectations. At least in one matter, clients’ expectations about rehabilitation outcomes were achieved: it is the case of clinical improvements, in terms of conditions with nursing needs. Otherwise, the overall good results―in terms of clinical, functional and
Table 2. Functional abilities (overall basic activities of daily living and walking) and conditions with nursing needs, at admission and at discharge.
social outcomes―overcame and covered other matching between expectations and outcomes. The topic of the association between improvements in walking ability with expectations not only in functional improvement, but also with clinical and overall improvement, is emblematic.
Our study has some limitations, including the focus in a unique sample, that hampers generalizability. More, McPhail and Haines warned about the risk of a self-fulfilling prophecy in case patients feeling helpless and anticipating poor levels of physical functioning, pain and depression might be less likely to participate in therapies and other treatments. Vice versa, patients exhibiting a positive outlook and high levels of self-efficacy might maximize their rehabilitation outcome through active participation during their rehabilitation stay  . A further weakness of our study is the lack of information about the quality of patients’ participation to the rehabilitative sessions. Yet, finding no quantitative differences in the attendance to rehab sessions might mitigate the risk of such self-fulfilling prophecy. The study has also some strengths. It is in line with some previous experience. In an Australian study, old patients admitted for subacute in-hospital rehabilitation were able to anticipate their discharge health-related quality of life with a moderate level of accuracy. Such ability held also for patients with poorer cognition  . Patients’ views might be mirrored by professionals’ ones, according to some literature hints. Physiotherapists in the same Australian subacute rehabilitation setting predicted their patients’ discharge health-related quality of life with substantial accuracy  . The prediction accuracy was higher among patients with better cognition than patients with poorer cognition.
Expectations declared by our patients are similar to those disclosed in a rehabilitative sample studied in the frame of goal attainment  .
Previously, we shown elsewhere that customer satisfaction was realistically related to reasonably attainable functional as well as clinical outcomes, in the same rehabilitative setting  . Further, we found that trends in self-perceived health paralleled improvements in objective functional gauges and subjective indicators, as expected  . Yet our correlations were weak, according to the literature  .
Poor outcomes of patients short of expectations raise concern. In wealthy persons, positive expectations regarding aging were associated with better physical function while getting old, as well as with increased longevity. Vice versa, those with more negative expectations were less likely to report engaging in physical activity  . The issue of keeping the locus of control maintains its basic importance for elderly individuals with somatic health problems   .
In a geriatric rehabilitative setting, most dependent, comorbid, frail elderly patients stated a set of reasonable expectations about their outcomes. Improvement in actual outcomes matched expected improvements in two circumstances: recovery in walking ability and in conditions with nursing needs. The high yield of different favorable outcomes may have masked the statistical significance for correspondence between actual and expected results. Special attention must be paid towards patients short of expectations, as their actual outcomes result poorer.
We are grateful to the whole staff of the ward for the commitment to patients’ expectations and outcomes, and to ASP administration for the support in our effort.