Hospital pneumonia and the toothbrush: the cheapest intervention nobody bills for
### The evidence on oral care preventing hospital-acquired pneumonia is real but uneven, and that unevenness is itself a clue to how the system prices the mouth.
Non-ventilator hospital-acquired pneumonia, NV-HAP, pneumonia that develops in hospitalized patients who are not on a ventilator, is one of the most common healthcare-associated infections in the United States, and one of the least discussed. A national analysis by Giuliano, Baker, and Quinn found an NV-HAP incidence of 1.6%, or 3.63 cases per 1,000 patient-days, with NV-HAP patients showing a greater likelihood of death than most other hospital-acquired infection comparison groups [1]. The intervention with the most consistent signal for preventing it is also one of the least resourced: brushing patients' teeth.
What the evidence actually shows
The honest starting point is a 2020 systematic review and meta-analysis by Satheeshkumar, Papatheodorou, and Sonis in the British Dental Journal, which pooled randomized and non-randomized studies of oral care interventions against NV-HAP. Restricted to the six randomized controlled trials (3,891 patients), the pooled result did not reach statistical significance: a risk ratio of 0.89 (95% CI 0.64-1.25, p=0.50). Non-randomized trials likewise showed no significant effect. But a subgroup analysis isolating studies where a dental professional was involved in oral care delivery, rather than oral care delivered generically by nursing staff, found a statistically significant risk reduction: RR 0.65 (95% CI 0.43-0.98, p=0.03) [2]. Read plainly, this meta-analysis does not support a blanket claim that "oral care prevents pneumonia" from the RCT evidence alone; it supports a narrower and more interesting claim, that who delivers oral care and how it's structured may matter as much as whether it happens at all.
The 2018 AHRQ "Making Healthcare Safer IV" evidence review reached a similarly disciplined conclusion: across the oral care protocol studies it assessed, the strength of evidence for NV-HAP incidence reduction was rated "Insufficient," due to inconsistent study designs, lack of standardized definitions of what counted as an NV-HAP case, and other methodological limitations, even though individual studies within that body of evidence showed effects, including one showing mortality dropping from 38.5% to 15.4% [3]. AHRQ's conclusion was not that oral care doesn't work. It was that the existing evidence base couldn't yet support a confident population-level estimate of how much it works.
Since those reviews, several implementation studies have reported larger, more specific effects. A four-unit cluster-randomized study found that increasing oral care frequency from roughly once to about twice daily was associated with an 85% reduction in NV-HAP incidence rate on the intervention units [4]. A community hospital quality-improvement initiative (Coury and Dietz, 2022) reported an oral care bundle, new oral care devices, staff education, and daily compliance audits, cutting the NV-HAP rate from 1.2 to 0.54 per 1,000 patient-days (a 58% reduction), the NV-HAP-attributed sepsis rate by 41%, and NV-HAP-attributed mortality by 50% [5]. Most recently, a multi-centre stepped-wedge cluster-randomized trial across nine wards in three Australian hospitals (over 8,000 patients, results presented at ESCMID Global 2026) reported NV-HAP incidence falling from 1.00 to 0.41 cases per 100 admission-days, roughly a 60% reduction, alongside oral care provision rates rising from 15.9% to 61.5% of eligible patient-days [6]. That trial is, as of this writing, reported from conference presentation; full peer-reviewed publication detail, including confidence intervals, was not available at the time of this brief and should be checked before being cited as a settled RCT result.
Taken together, the pattern across these studies is consistent even where the meta-analytic pooling is not: when oral care actually happens, consistently and as a structured protocol rather than an occasional gesture, NV-HAP rates fall in nearly every study that measured it, and the size of the effect tracks with how rigorously the protocol was implemented and audited. The 2020 meta-analysis's null RCT-pooled result is best read as a comment on inconsistent implementation across older trials, not as evidence that oral hygiene has no relationship to pneumonia risk, a relationship that is biologically well-established: dental plaque is a reservoir for respiratory pathogens, and reducing that reservoir reduces the aspiration risk pathway into the lungs.
Why it's optional in most hospitals
If the intervention is this cheap, a toothbrush, toothpaste, and a few minutes of nursing time, the obvious question is why it isn't universal. AHRQ's review names the barriers directly: oral care is widely perceived by staff as an optional comfort measure rather than a clinical infection-prevention task, hospitals frequently supply inadequate oral care materials, high patient-to-staff ratios and time pressure crowd it out, and, critically, hospitals are not required to monitor NV-HAP incidence at all, unlike ventilator-associated pneumonia, which has mandatory surveillance and public reporting obligations that make it a tracked quality metric [3]. What isn't measured isn't managed, and what isn't billed isn't prioritized within a system organized around reimbursable procedures. Oral care by nursing staff is not a separately billable line item; the infection it may prevent, extended length of stay, sepsis workup, ICU transfer, is billable, but by the time those costs appear, the causal thread back to a missed teeth-brushing shift is invisible in the chart.
What this says about how the system prices the mouth
This is the pattern that recurs across every clinical setting where the mouth intersects with the rest of medicine: the connective tissue between oral health and systemic outcomes is well enough established to write into cancer protocols, cardiac surgery clearance forms, and hospital infection-prevention literature, and poorly enough resourced that the actual bedside intervention, a toothbrush, a trained hand, a monitored protocol, remains optional, unstandardized, and unbilled. NV-HAP is not obscure. It is common, costly, and its cheapest countermeasure sits unfunded next to interventions that cost orders of magnitude more.
What this means in practice
If you have a hospitalized family member who is not eating or drinking normally, ask directly whether oral care is part of their nursing care plan and how often it happens, because in many hospitals the honest answer is "when someone remembers." For dental practices and hospital systems building or auditing oral care protocols, the evidence supports structured, audited, professionally-informed programs over ad hoc practice; the clearance and protocol standard used on the pre-surgical side of this same oral-systemic link is documented at predentalcheck.com, with the physician and care-coordinator side at premedicalcheck.com.