Answers are grounded in this paper and cite their sources.
According to an African proverb, ‘there are two things over which you have complete dominion, authority and control—your mind and your mouth’. Geriatricians are fully aware that in many of their patients the first half of the proverb may, at least temporarily, not apply. The current edition of Age and Ageing contains two original contributions which, especially when taken in conjunction, may challenge our perception and practice relating to the second half, and in particular to prevention of upper respiratory infection in the frail hospitalised or institutionalised older person. In the first contribution, Ewan and colleagues [1] demonstrate that in a convenience sample of 29 cognitively intact orthopaedic inpatients (24 women), with a mean age of 81.5 years, bacteria known to be responsible for hospital-acquired pneumonia can be isolated from the mouth/oropharynx (hereafter abbreviated as ‘mouth’) of ~45% of subjects. Culture of a simple 20-s oral rinse with sterile water seems to be the most sensitive method of detection. In the second contribution, Cabre and colleagues [2] showed in a prospective study that, in 134 patients (80 men) with a mean age of 84.5 years, consecutively admitted to an acute geriatrics unit with well-documented pneumonia, 55% had clinical signs of dysphagia on bedside assessment using well-validated and discriminative methodology. The patients in Cabre’s study, though pre-selected for a diagnosis of pneumonia, were otherwise representative of the clinical spectrum of hospitalised acutely ill elderly people, with multiple co-morbidity and high levels of both functional and cognitive impairment. In assessing the implications of these findings, several questions arise. (i) Have the above findings any supportive evidence from other authors? (ii) Should we be instituting a policy of universal screening for dysphagia in elderly people presenting with pneumonia (or indeed in the frail, hospitalised elderly in general)? (iii) Should we modify our treatment of pneumonia in older people to ‘cover’ Gram-negatives and staphylococci? (iv) What evidence is there that mouth colonization by potentially unpleasant bacteria is of any clinical significance? (v) Should we screen for mouth colonization? (vi) Can we eradicate such colonization, and if so does this affect the incidence or prevalence of morbidity and/or mortality from lower respiratory tract infection (LRTI)? (vii) If eradication attempts are deemed useful and feasible, what methods should we use? (viii) What are the possible side effects of such policies and what will be the costs? Let us examine the mouth colonization issues first. The high prevalence of mouth colonization by Gram-negative bacilli in (often) elderly, frail, functionally dependent, multiply co-morbid people in hospital, long-term care and (even) community settings by Gram-negative bacilli is well established. Prevalence may be of the order of 15–50%, depending upon the sample chosen [3–10], and may reflect illness severity [3, 8–10]. Colonization is associated with increased rates of infection (including LRTI) and with medium term mortality in older people [7], and the US Centers for Disease Control (CDC) recognises aspiration of oropharyngeal organisms as an important aetiological route for the development of nosocomial pneumonia [11]. Attempts to eradicate oral carriage or reduce the oral burden of Gram-negative flora in patients including the frail, elderly institutionalised population have varied in intensity. A study at one end of the intervention spectrum employed a complex regime of nursing and/or caregiver tooth and mucosal brushing for 5 min after each meal, together with weekly oral hygiene visits from a dentist or dental nurse [12] (in dependent, often cognitively impaired elderly nursing home residents). This regime would necessitate ~44 h of carer/nurse time per week in a rest home or ward with 25 residents/patients. The other end of the spectrum in terms of resource allocation was a study employing the simple use of chlorhexidine mouthwash in young adult patients prior to cardiac surgery [13]. Outcome measures in these studies have included febrile days, total respiratory infection rates, nosocomial infection rate, pneumonia rate and death from pneumonia, with statistically and (highly) clinically significant reductions in all adverse outcomes [12, 13]. As a result of these observations, the CDC recommends the development and implementation of a ‘comprehensive oral-hygiene program (that might include the use of an antiseptic agent) for patients in acute-care settings or residents in long-term-care facilities who are at high risk for health-care-associated pneumonia’ [11]. The ‘simple’ use of antiseptic mouthwash may of course be somewhat less than simple in, for example, cognitively impaired patients who may not cooperate or may swallow the antiseptic. The same arguments, however, apply to screening for colonization with oral sterile water rinse, and it would thus seem sensible to adopt a pragmatic approach tailored to the needs and abilities of individual residents/patients. This is of course the accepted practice in many if not all of the health care settings in which the readers of this journal operate. The question that arises is how many of us, even in the context of nosocomial LRTI in our patients, take the trouble to encourage rigorous compliance with such practices and to investigate why they might have failed to prevent LRTI in individual cases. The current contribution by Cabre and colleagues [2] as well as recent literature supports the view that aspiration is much more common than was once believed. Cabre’s patients were selected for a diagnosis of pneumonia, and other studies have reported very high prevalences of dysphagia in patients with stroke and with degenerative neurological disease. However, clinical features suggesting aspiration have also recently been found in 51% of residents of Taiwanese long-term care facilities (supporting previous data in this area) [14], and studies of the ‘symptoms’ of dysphagia (though not necessarily proven aspiration) report prevalences of between 14% and 33% of community-dwelling older people [15–18]. The risk factors for aspiration in Cabre’s study [2] are not surprising and reflect the literature: older age, greater functional impairment, prior institutionalization, sedative use, stroke, dementia and a high prevalence of the ‘geriatric giants’ and other common geriatric syndromes (e.g. pressure sores). However, the presence of aspiration was also a strong predictor of pneumonia severity and of in-hospital mortality, new institutionalization and mortality at both 1 month and at 1 year. Such findings must at the very least promote debate about current treatment regimes for community acquired pneumonia in older people (especially those with any of the risk factors listed), as well as the possible benefits and dis-benefits of more widespread dysphagia assessment. The staffing resources needed for widespread ‘formal assessment’ of all those deemed at risk (largely speech and language therapists (SALTs) but also training of medical and nursing staff both in awareness and in screening techniques—given that SALTs generally do not provide routine out of hours cover) would be considerable. However, as a ‘screening test’ (as opposed to a detailed formal assessment), the use of the combination of a water swallow test and measurement of oxygen desaturation, taking 5–10 min to perform in most cases, has, in stroke patients for example, been shown to have a positive predictive value and specificity approaching 100%, and sensitivity and negative predictive values of 65–70% compared to the gold standard of videofluoroscopy [19]. It would be important to ensure that the detection of a large proportion of ‘abnormal swallows’ did not result in an overzealous application of, for example, modified consistency diets which patients found unpalatable—with potential adverse consequences for hydration and nutrition, and for mood. Such consequences could of course in turn lead to a legion of important secondary adverse outcomes. If admission screening of acute hospital admissions in at-risk individuals were to become more widespread, it would also be essential to ensure that a reassessment (using more resources) was performed before discharge in order to detect (not unlikely) improvement and thus not indefinitely consign a patient to a modified diet. All these considerations strongly suggest the need for a randomised controlled trial (or trials of individual aspects of the above) which given the complexity of the situation would need to be sufficiently large to be adequately powered. The ‘treatment’ of pneumonia in older people, whether acquired in the community, in institutional care or in hospital, is beyond the scope of this editorial. However, the current contributions by Ewan and colleagues [1] and by Cabre and colleagues [2] must at the very least promote a higher and wider index of suspicion around aspiration of pathogenic, largely Gram-negative bacteria, with an inclination to the use of antibiotics appropriate for such pathogens as an (early) second line. There is no conflict of interest to declare.
No takes yet. Share an insight, caveat, or question.
Martin J. Connolly (2009) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: