Who Died of Mpox in the United States, 2022–2023

This page summarizes a report in CDC's MMWR by CDC and state and local health departments, covering May 10, 2022 to March 7, 2023. The report was later corrected in a table and its acknowledgments; the text summarized here was not affected.

As of March 7, 2023, the United States had reported 30,235 confirmed and probable mpox cases, mostly among cisgender men who reported recent sexual contact with another man. Most cases in the outbreak cleared up on their own, but some people became severely ill and some died, particularly those whose immune systems were moderately to severely compromised.

How deaths were counted

Health departments reported cases and deaths through national case surveillance, and CDC clinical officers gathered more detail on some patients through consultations with their clinicians. A death counted as mpox-associated if mpox appeared on the death certificate, either in the chain of events that directly caused death (Part I) or among significant contributing conditions (Part II). Decedents were compared with patients presumed to have survived.

That is 1.3 mpox-associated deaths per 1,000 cases, against about 1.2 per 1,000 worldwide. Twenty-five of the 38 deaths (65.8%) came in October–November 2022.

Who died

• Nine of 10 decedents with information on recent sexual or intimate contact had had it with cisgender men in the 3 weeks before symptoms began. Two reported close nonsexual contact with people with mpox, such as co-sleeping or caring for a household member.
• Five of 11 with information were experiencing homelessness.

Their illness and treatment

Among the 27 decedents with clinical data, the median time from first symptoms to death was 68 days (range 1–146), and 20 of 23 with information went to an intensive care unit.

Mpox treatments. Twenty-five of 27 received mpox-directed therapy at some point:

Of the 25 given tecovirimat, 15 (60.0%) got it within 3 days of diagnosis, but six (24.0%) not until 3 weeks or more later. All 24 decedents with data had lesions described as necrotic, diffuse, or worse after a 14-day tecovirimat course.

• Two received no mpox therapy, at clinicians' discretion. One had other conditions that raised concerns about contraindications. The other was on antiretroviral therapy (ART) for HIV with an undetectable viral load when first assessed, and was found dead at a wellness check about a month later with diffuse lesions characteristic of mpox.
• Seven of 27 refused treatment or intravenous medicines, or left the hospital against medical advice, at some point.
• More than half of 24 with information (54.2%) received steroids, for mpox complications or for concern about immune reconstitution inflammatory syndrome (IRIS), a hyperinflammatory response that can occur in people with HIV during the first 6 months of ART.

HIV and other immune problems. Thirty-one of 33 decedents with complete data had HIV. Of 24 with CD4 data, every one had a count below 200 cells/mm³, and 23 below 50. Of the two immunocompromised decedents without HIV, one was presumed immunocompromised by undiagnosed diabetes and had diabetic ketoacidosis when mpox was diagnosed; the other was severely immunocompromised after a recent renal transplant complicated by acute rejection.

Only 2 of 25 decedents with HIV had been taking ART before their mpox diagnosis, and in one of them HIV was poorly controlled. ART was started for 19 of the 20 not on it, including one person diagnosed with HIV 5 days after mpox; one declined treatment for advanced HIV; status was unknown for three. For seven, clinicians delayed or interrupted ART out of concern about IRIS.

What it means

• Immune weakness drove the deaths. Nearly all decedents were immunocompromised when diagnosed, and the long illnesses likely reflect a reduced ability to fight infection. Before 2022, severe mpox occurred mainly where the disease is endemic and resources are limited, and where tecovirimat, cidofovir and VIGIV are not routinely available.
• Treat early. Most decedents got prompt treatment and intensive care, but nearly a quarter waited 3–7 weeks between diagnosis and treatment. When mpox is suspected, clinicians should consider early mpox-directed therapy, especially for immunocompromised patients.
• Start ART; avoid immune suppression. Boosting immune function with ART and avoiding immunosuppressive drugs such as steroids are critical to recovery. So far, no evidence shows that IRIS worsens mpox outcomes. People with mpox and HIV who are not on ART, including those newly diagnosed, should start it as soon as possible; HIV-negative people should be assessed for pre-exposure prophylaxis.
• Test for HIV. Every patient with suspected mpox should be checked for immunocompromising conditions such as HIV, and offered HIV testing.
• Disparities. Almost 90% of deaths were in Black men, while fewer than one in three survivors were Black men. This matches earlier reports that most patients hospitalized with severe mpox were Black men with uncontrolled HIV, and mirrors disparities in HIV itself: in 2020, 75% of all-cause deaths among adults with HIV were in males, 39% of them Black males. Barriers exist at every level of HIV care, from recognizing risk to getting tested, pre-exposure prophylaxis and ART.
• Psychosocial support. Studies of mpox patients in Nigeria documented anxiety, depression and suicide, possibly linked to stigma, to the implications of a sexually associated infection, and to isolation during long treatment. Black and Hispanic or Latino men in the United States may also face language barriers, homophobia and discrimination. One of the non-mpox-associated deaths was a suicide, and one decedent persistently declined HIV treatment, pointing to a need for stronger psychosocial support in the response.

The report calls for integrating prevention, testing and treatment for sexually associated infections such as mpox and HIV, and for equitable access to care for both, particularly for Black men and others at risk.

Limitations

• Deaths may be undercounted: reporting guidance was written during the outbreak, delays were expected, and clinical consult reports were gathered passively.
• Some data, such as housing, treatment, HIV status and CD4 count, were more often available for decedents than survivors, because patients receiving intensive mpox treatment had extra reporting requirements. That limits comparisons between the groups.

Sources

• Riser AP, Hanley A, Cima M, et al. "Epidemiologic and Clinical Features of Mpox-Associated Deaths — United States, May 10, 2022–March 7, 2023." MMWR 72(15). https://www.cdc.gov/mmwr/volumes/72/wr/mm7215a5.htm
• Erratum, MMWR 72(31), correcting Table 1 (smallpox vaccination) and the acknowledgments: https://www.cdc.gov/mmwr/volumes/72/wr/mm7231a5.htm
• The report's figure is not reproduced, because the report was prepared jointly by CDC and state and local health departments.