The long-term prognosis of patients with delirium in the acute phase of stroke: PRospective Observational POLIsh Study (PROPOLIS)

Delirium is a common complication of stroke, with the frequency estimated to be between 10 and 48% [1]. There is abundant data linking delirium to adverse clinical outcomes in medical and surgical geriatric populations; however, post-stroke delirium remains understudied with little known about its effect on clinical outcome and long-term prognosis. Additionally, the limited available studies on long-term prognosis for patients with delirium in the acute phase of stroke show inconsistent data regarding long-term life expectancy [2‐8].

Long-term medical consequences for both delirious patients and healthcare budgets are serious. Considering the inconsistent data from previous studies and the lack of data on the outcome of delirium in the acute phase of stroke in the Polish population, the aim of this single-center PRospective Observational POLIsh Study on post-stroke delirium (PROPOLIS) was to determine the prognostic significance of delirium for discharge destination, recurrence of stroke, cardiovascular events, functional status, nursing home admission, and death. This was assessed by following a large cohort of patients with stroke for 12 months.

Of the 750 patients included 682 stroke survivors were scheduled for follow-up visits (a flow chart shows the study design). From them, 67 died and 43 were lost from the follow-up 3 months post-stroke. Ultimately, patients with delirium in the acute phase of stroke experienced increased mortality within 3 months post-stroke (26.62% vs. 5.36%, Person’s χ² 56.302 p < 0.001; survival comparison shown in Fig. 2—Gehan’s Wilcoxon test p < 0.001). The baseline characteristic of the stroke population and risk factors for death after 3 months are shown in Table 1. Patients who died were also significantly older, had more sever delirium and cognitive impairment prior to and during hospital stay, had more sever concomitant chronic diseases (in particular atrial fibrillation and diseases affecting respiratory tract) and higher temperature between days 1 and 7 of hospitalization. Death was also significantly more common in those with aphasia, neglect and vision deficit and in those more disabled on admission and on discharge.

Delirium was found to be an independent risk factor for death in univariate logistic regression analysis (OR 6.41 CI 3.76–10.92; p < 0.001). In the multivariable logistic regression model, delirium remained significantly associated with mortality 3 months after stroke (results are shown in Table 2).

Twelve months after their strokes, 485 patients attended follow-up visits (52 died and 35 were lost from the follow-up). Mortality 12 months post-stroke was higher in those with delirium (42.66% vs. 12.58%, Person’s χ² 62.406; p < 0.001; survival comparison shown in Fig. 2—Gehan’s Wilcoxon test p < 0.001). Patients who died were also older, had more severe delirium and cognitive impairment prior to and during hospital stay, were less educated, had longer hospitalization, more severe concomitant chronic diseases (in particular atrial fibrillation and diseases affecting respiratory tract) and higher temperature between 1 and 7 day of hospitalization. Death was also more often in those with aphasia, vision deficit and in those more disabled on admission and on discharge.

Delirium was found to be an independent risk factor for 12 months mortality in univariate logistic regression analysis (OR 5.17 CI 3.36–7.96; p < 0.001). In multivariable logistic regression model, delirium remained significantly associated with 12-months post-stroke mortality (results are shown in Table 2).

Patients with delirium were also more disabled and reported a larger impairment on daily activities than those without delirium 3 and 12 months post stroke (Table 3 shows these results).

We did not find a significant association between delirium and recurrence of stroke or cardiovascular complications 3 and 12 months after stroke.

The results of this study showed that delirium is independently associated with a poorer prognosis both 3 and 12 months after stroke, including greater functional disability, higher rates of admission to long-term institutionalized facilities, and greater mortality, when compared to delirium-free patients.

This study identified a number of risk factors for post-stroke mortality. Within these factors, increased temperature in acute phase of stroke and delirium were the only modifiable risk factors.

Our study confirmed some but not all of previously reported results. In a study by Dostovic et al. [4] that followed 100 patients with delirium and 100 controls with stroke, delirious patients had a significantly higher mortality and a greater degree of functional disability in the first year after ischemic stroke. Sheng et al. [2] followed 140 patients with stroke and found that delirium predicted 6 and 12 months mortality, disability and institutionalization. Both Henon et al. [6], who followed 202 patients with stroke, and McManus et al. [5], who recruited 82 patients, did not find increased mortality at 6 and 12 months, respectively. Miu et al. [3] included 314 patients with stroke and followed them for 1 year. This study found that patients with delirium had more functional disability and more often were placed in nursing homes at discharge. Mortality 6 and 12 months after stroke was higher in patients with delirium but the differences were not significant. Oldenbeuving et al. [7] who followed 527 patients with stroke found that delirium was associated with lower functional outcome and higher mortality 1-month post stroke. Qu et al. [8] who observed 261 patients with ischemic stroke for 6 months found that delirium was associated with worse functional outcome after 3 months.

Stroke patients often suffer from temperature elevations without an identifiable infection; possibly endogenous fever due to direct stroke effects on the brain [24]. We identified that increased temperature between first and third-day of the hospital stay independently increased mortality 3 months after stroke. The data about temperature management in acute stroke are inconclusive [25]. Our data provide support for trials testing temperature lowering in acute phase of stroke as a preventing strategy for unfavorable long-term outcome.

Results of our study confirm the evidence from medical literature that delirium predicts a poor outcome [26]. Previous studies on the long-term outcome of post-stroke delirium were conducted on small cohorts. The large number of consecutive patients with stroke is a strength of our study—there is no other prospective study on post-stroke delirium with such a big cohort. The other strength of our study was the constant observation of patients by medical personnel and careful, formal assessment of delirium conducted every day. For delirium screening we used bCAM for verbal patients or CAM-ICU for patients who could not speak but were able to communicate in a non-verbal way. Both methods have high sensitivity and specificity and are easy to administer [10, 11]. The same assessor administered the scale from day 1 to day 7, thus making the bias of inter-observer variation minimal. All this allowed for the identification of all delirium cases, even those lasting briefly.

This study included patients of a wide age rage. Enrollment of younger patients might decrease the percentage of cases with post-stroke delirium. However, the mean age of the cohort was high and similar to other studies, therefore, we do not think that including younger patients would cause a bias. The number of patients lost from follow-up was small.

A limitation of our study is that the incidence of delirium in the acute phase of stroke in our sample might be underestimated due to an observation period restricted to only 7 days. This is the median duration of hospital-stay in Krakow’s stroke unit. Therefore, those with delayed onset delirium were missed.

Delirium is a common complication of stroke and its long-term consequences are very serious. Predisposing factors for its development were previously identified for the Polish population [7]. Unfavorable outcomes of patients with delirium in the acute phase of stroke emphasize the need of better awareness of this common stroke complication. As a matter of its course, data shows delirium is strongly associated with additional healthcare costs [27]. Bearing this in mind, results of our study support the need for further studies addressing the influence of early recognition and treatment of delirium in the acute phase of stroke as well as testing lowering temperature to decrease later post-stroke mortality and to decrease adverse long-term outcomes within this population, also to better appraise the associated healthcare burden.