Clozapine and Pneumonia: Infection Risk, Clozapine Levels, and Dose During Illness

Creator: Ryan S. Sultan, MD
Published: 2026-11-12

By Ryan S. Sultan, MD
Psychiatrist · Columbia University Irving Medical Center
Published November 12, 2026 · 6-minute read

In a Finnish register study of 2,659 people with schizophrenia who took clozapine, the cumulative incidence of pneumonia was 29.5% at 20 years after starting, and pneumonia was associated with increased mortality among those clozapine users (odds ratio 2.8).¹ The abstract gives no pneumonia figure for people not taking clozapine and no comparison group for the odds ratio.

Fever or signs of infection on clozapine mean contacting the prescriber the same day. A blood count may be needed, and the label directs prescribers to evaluate fever carefully to rule out severe neutropenia or infection.² Any dose change during an illness is the prescriber's decision.

How much higher the risk is, and compared with what

Clozapine and infection risk in three national or territory-wide studies
Study	Outcome and comparison	Estimate
Finland, 61,889 people with schizophrenia	Hospitalization for pneumonia, compared with no antipsychotic use	Adjusted hazard ratio 1.44 for high-dose clozapine (330 mg per day or more) and 1.43 for medium-dose (180 mg to under 330 mg per day)
Taiwan, 33,024 inpatients with schizophrenia	Pneumonia requiring hospitalization, current clozapine use compared with no clozapine use, in a nested case-control analysis of 1,741 cases	Adjusted risk ratio 3.18 for current clozapine use
Hong Kong, 1,450 on clozapine and 9,601 on olanzapine	Any infection, clozapine compared with olanzapine	Weighted hazard ratio 1.25; 7.26 compared with 6.00 infections per 100 person-years

Sources: ^{3, 4, 5} All three are observational.

In the Finnish cohort, 8,917 patients (14.4%) had at least one pneumonia hospitalization over 22 years of follow-up, and 1,137 of them (12.8%) died within 30 days of admission. Compared with no antipsychotic use, antipsychotic use overall was not associated with pneumonia (adjusted hazard ratio 1.12, 95% CI 0.99 to 1.26). High-dose quetiapine (1.78) and high-dose olanzapine (1.29) were associated with increased risk on the same comparison.³

The Taiwanese authors described the clozapine association as dose-dependent.⁴ The Hong Kong study limited both groups to patients with treatment resistance, and respiratory and gastrointestinal infections were the main contributors to the difference.⁵

Why pneumonia: what the sources state

A 2020 consensus statement attributed the association to drooling and aspiration: clozapine "may be associated with a higher risk of pneumonia, likely due to sialorrhea and aspiration rather than neutropenia."⁶ A single-hospital study suggested immunologic factors or sedation and drooling that make aspiration more likely, and added that causative mechanisms require further investigation.⁷ The label lists aspiration and pneumonia among postmarketing reports.²

In the Finnish cohort, antipsychotics with a high anticholinergic burden were the only burden category associated with pneumonia (adjusted hazard ratio 1.26, compared with no antipsychotic use).³

Lower antibody levels are a second hypothesis. Immunoglobulins were reduced in 123 clozapine-treated patients compared with 111 who had never taken it.⁸ A 2026 systematic review of 16 studies concluded that no immune alteration can be confidently ascribed to clozapine independent of treatment-resistant illness.⁹ Drooling itself is covered on the clozapine drooling page.

Infection may raise the clozapine level

The label states that published case reports describe examples where pneumonia or other inflammatory conditions may increase clozapine concentrations. It adds that the clinical significance and the mechanism have not been fully characterized and may involve reduced CYP1A2 activity. The label has no instruction to lower the dose during infection.²

In a Norwegian drug-monitoring study of 126 patients, the dose-adjusted clozapine concentration correlated with CRP in non-smokers (R = 0.492) and showed no significant correlation in smokers.¹⁰ The page on clozapine levels covers how levels are measured and read.

Lowering the dose during infection is expert opinion

The 2020 consensus statement, written for the COVID-19 pandemic, called for an urgent physician assessment with a complete blood count for any symptoms of infection. With fever and flu-like symptoms, it said, "the emergence of signs and symptoms of clozapine toxicity may require clinicians to reduce the dose of clozapine by as much as half." It advised continuing the lower dose until 3 days after the fever has subsided and then increasing stepwise.⁶

An expert review from the same year goes further and advises cutting the dose in half once fever or pneumonia develops.¹¹ Both are instructions to clinicians, and a PubMed search on November 12, 2026 found no trial of dose reduction during infection. The decision to lower, hold, or continue clozapine during an illness belongs to the prescriber, and a patient or family member who notices a fever calls that day.

The label also tells prescribers to interrupt clozapine and obtain an ANC for a temperature of 101.3°F (38.5°C) or higher.²

COVID-19: findings in both directions

In a UK cohort of 6,309 people followed from March to May 2020, clozapine was associated with a higher risk of COVID-19 infection than other antipsychotics. The hazard ratio was 2.62 unadjusted and 1.76 after adjustment for confounders that included clinical contact.¹² A later study by the same group, of patients already infected, found no evidence of increased hospitalisation, intensive care treatment, or death with clozapine.¹³

A Stockholm register study of 8,233 people compared clozapine with other antipsychotics and found adjusted hazard ratios of 0.96 for inpatient care, 1.69 for intensive care, and 0.86 for death, all with confidence intervals that included no difference.¹⁴ A 2023 systematic review of 18 studies found no strong evidence of increased risk after adjustment for medical comorbidities.¹⁵ A 2026 meta-analysis pooled an odds ratio of 1.53 (95% CI 1.02 to 2.30) for infection with clozapine exposure (its abstract names no comparison group) and reported high heterogeneity (I² = 91.2%).¹⁶

Vaccination advice comes from reviews

The label contains no vaccination advice.² Published recommendations come from reviews: one called for prioritising clozapine users for all vaccinations, particularly pneumococcal vaccination, and a Dutch group called for early COVID-19 vaccination.^{17, 18} A systematic review of 14 records (5 original articles and 9 case reports) found no evidence that clozapine impairs the effectiveness of vaccines and no serious safety concerns, and described the data as limited. Two of its studies indicated that clozapine use may be associated with reduced immunoglobulin levels.¹⁹

Limitations of the evidence

This evidence has limitations. First, every estimate is observational, and confounding by treatment-resistant illness can't be excluded. Second, the dose-reduction and vaccination advice rests on case reports, reviews, and expert consensus. Nevertheless, the association has appeared in data from Finland, Taiwan, and Hong Kong, and fever, cough, or shortness of breath on clozapine warrants a call to the prescriber the same day.

Frequently asked questions

Does clozapine increase the risk of pneumonia?

Observational studies associate clozapine with pneumonia. In a Finnish register study of 2,659 people taking clozapine, the cumulative incidence of pneumonia was 29.5% at 20 years, and in a second Finnish cohort medium-dose and high-dose clozapine were associated with pneumonia hospitalization (adjusted hazard ratios 1.43 and 1.44) compared with no antipsychotic use. These studies can't establish cause.^{1, 3}

Should clozapine be reduced when you're sick?

That is the prescriber's decision, and it starts with a same-day call. A 2020 consensus statement says signs of clozapine toxicity during fever and flu-like symptoms may require clinicians to reduce the dose by as much as half. The label has no instruction to lower the dose during infection, and a PubMed search on November 12, 2026 found no trial of dose reduction during infection.^{6, 2}

What should I do if I get a fever on clozapine?

Contact the prescriber the same day. A blood count may be needed, and the label tells prescribers to interrupt clozapine and obtain an ANC for a temperature of 101.3°F (38.5°C) or higher. Any change to the dose is the prescriber's decision.²

Does clozapine raise the risk of COVID-19?

The findings conflict. A UK cohort of 6,309 people found a higher risk of infection with clozapine than with other antipsychotics early in the pandemic (adjusted hazard ratio 1.76), a later study of infected patients found no evidence of increased hospitalisation, intensive care treatment, or death, and a 2026 meta-analysis pooled an odds ratio of 1.53 for infection with clozapine exposure (its abstract names no comparison group), with high heterogeneity (I² = 91.2%).^{12, 13, 16}

Should people on clozapine get the pneumonia vaccine?

The label contains no vaccination advice, and the published recommendations come from reviews. One review called for prioritising clozapine users for all vaccinations, particularly pneumococcal vaccination. A systematic review of 14 records found no evidence that clozapine impairs the effectiveness of vaccines and described the data as limited.^{2, 17, 19}

In a crisis: call or text 988, or go to the nearest emergency department. A first episode of psychosis belongs in a hospital-based or coordinated specialty care program.

I start and manage clozapine in an outpatient practice in Chelsea, Manhattan. Details are on the schizophrenia care in NYC page. Clozapine should not be stopped, skipped, or restarted without the prescriber. This page is general information and does not replace an evaluation.