Stroke-associated pneumonia: major advances and obstacles.

BACKGROUND Stroke-associated pneumonia (SAP) has been implicated in the morbidity, mortality and increased medical cost after acute ischemic stroke. The annual cost of SAP during hospitalization in the United States approaches USD 459 million. The incidence and prognosis of SAP among intensive care unit (ICU) patients have not been thoroughly investigated. We reviewed the pathophysiology, microbiology, incidence, risk factors, outcomes and prophylaxis of SAP with special attention to ICU studies. METHODS To determine the incidence, risk factors and prognosis of acute SAP, PubMed was searched using the terms 'pneumonia' AND 'neurology intensive unit' and the MeSH terms 'stroke' AND 'pneumonia'. Non-English literature, case reports and chronic SAP studies were excluded. Studies were classified into 5 categories according to the setting they were performed in: neurological intensive care units (NICUs), medical intensive care units (MICUs), stroke units, mixed studies combining more than one setting or when the settings were not specified and rehabilitation studies. RESULTS The incidences of SAP in the following settings were: NICUs 4.1-56.6%, MICUs 17-50%, stroke units 3.9-44%, mixed studies 3.9-23.8% and rehabilitation 3.2-11%. The majority of NICU and MICU studies were heterogeneous including different neurovascular diseases, which partly explains the wide range of SAP incidence. The higher incidence in the majority of ICU studies compared to stroke units or acute floor studies is likely explained by the presence of mechanical ventilation, higher stroke severity causing higher rates of aspiration and stroke-induced immunodepression among ICU patients. The short-term mortality of SAP was increased among the mixed and stroke unit studies ranging between 10.1 and 37.3%. SAP was associated with worse functional outcome in the majority of stroke unit and floor studies. Mortality was less consistent among NICU and MICU studies. This difference could be due to the heterogeneity of ICU studies and the effect of small sample size or other independent risk factors for mortality such as the larger neurological deficit, mechanical ventilation, and age, which may simultaneously increase the risk of SAP and mortality confounding the outcomes of SAP itself. The pathophysiology of SAP is likely explained by aspiration combined with stroke-induced immunodepression through complex humeral and neural pathways that include the hypothalamic-pituitary-adrenal axis, parasympathetic and sympathetic systems. CONCLUSIONS A unified definition of SAP, strict inclusion criteria, and the presence of a long-term follow-up need to be applied to the future prospective studies to better identify the incidence and prognosis of SAP, especially among ICU patients.