In August 2017, Hurricane Harvey caused unprecedented flooding and devastation to the Houston metropolitan area (1). Mold exposure was a serious concern because investigations after Hurricanes Katrina and Rita (2005) had documented extensive mold growth in flood-damaged homes (2,3). Because mold exposure can cause serious illnesses known as invasive mold infections (4,5), and immunosuppressed persons are at high risk for these infections (6,7), several federal agencies recommend that immunosuppressed persons avoid mold-contaminated sites (8,9). To assess the extent of exposure to mold and flood-damaged areas among persons at high risk for invasive mold infections after Hurricane Harvey, CDC and Texas health officials conducted a survey among 103 immunosuppressed residents in Houston. Approximately half of the participants (50) engaged in cleanup of mold and water-damaged areas; these activities included heavy cleanup (23), such as removing furniture or removing drywall, or light cleanup (27), such as wiping down walls or retrieving personal items. Among immunosuppressed persons who performed heavy cleanup, 43% reported wearing a respirator, as did 8% who performed light cleanup. One participant reported wearing all personal protective equipment (PPE) recommended for otherwise healthy persons (i.e., respirator, boots, goggles, and gloves). Immunosuppressed residents who are at high risk for invasive mold infections were exposed to mold and flood-damaged areas after Hurricane Harvey; recommendations from health care providers to avoid exposure to mold and flood-damaged areas could mitigate the risk to immunosuppressed persons.

Interviews were conducted with a convenience sample of immunosuppressed residents from three hospital systems in the Houston metropolitan area. Eligible residents were selected because of risk factors for invasive mold infections (7); participants included persons who had received a solid organ transplant in the past year or who had been prescribed an immunosuppressive medication, including tumor necrosis factor inhibitors, cyclosporine, or chemotherapeutic agents, in the last 3 months. Models developed by CDC’s Geospatial Research, Analysis and Services Program were used to predict whether residents’ homes had been flooded. Residents whose homes were predicted to have been flooded were prioritized for contact.

Among the three hospital systems from which participants were selected, systematic, hospital-wide messaging about avoiding mold exposure had not been disseminated before Hurricane Harvey. A CDC questionnaire developed after Hurricanes Katrina and Rita was modified and field-tested. Questions were focused on experiences with housing, flooding and mold, cleanup activities, and PPE. Cleanup was categorized as either heavy (e.g., removing furniture, drywall, or carpeting) or light (e.g., sweeping, wiping off counters or walls, or retrieving personal items). The 20-minute questionnaire was administered by telephone in either English or Spanish by trained interviewers during October 21–November 8, 2017; no personally identifiable information was collected. Interviewers were trained on types of PPE. Federal agency–developed materials on how residents who are not immunosuppressed can safely conduct mold cleanup after disasters were made available to participants after the interview. Because this work was part of an ongoing public health response, it was determined to be nonresearch public health practice by CDC’s Human Research Protection Office and the local hospitals and thus was not subject to institutional review board review.

Interviewers attempted to contact 800 immunosuppressed persons, and 167 (21%) were reached, 109 (65%) of whom consented to be interviewed. Among these 109 persons, 103 (94%) had stayed within the Houston metropolitan area during Hurricane Harvey (August 25–August 29, 2017) or in the months afterwards (August 30–interview date); the survey sample consisted of these 103 persons. During the hurricane, 77 (75%) reported staying at home for the majority of the time, 20 (19%) stayed at friends or relatives’ homes, and seven (7%) stayed at other places, including hotels, hospitals, shelters, or nursing homes (Table 1). In contrast, since the hurricane, 83 (81%) stayed at home, 12 (12%) stayed at friends or relatives’ homes, and six (6%) stayed at other places. Of the 102 participants whose responses were available, 92 (89%) reported living in their homes at the time of the interview; of the 10 who did not, eight had been back to their homes. One participant’s response was not available.

Forty-six (45%) respondents reported that water had entered their homes, 37 (80%) of whom reported that the first floor living space was flooded with a median of 3 inches of water (interquartile range [IQR] = 1–12 inches) for a median of 3 days (IQR = 1–4 days) (Table 2). Among the 37 respondents who reported water in their living space, 28 (76%) reported seeing or smelling mold inside the home after the hurricane, and 32 (86%) had their home cleaned for water damage or remediated for mold. Seventeen (53%) participants lived in the house during cleanup, and 17 (53%) performed the cleanup themselves. In addition, 23 (62%) reported plans for cleanup or remediation within the next 6 months.

Participation in cleanup activities for any home was assessed among all 103 survey respondents; overall, 50 (49%) engaged in any cleanup activities, including 23 (22%) who engaged in heavy cleanup activities for a median of 7 days (IQR = 5–14 days) and 27 (26%) who participated in only light cleanup activities for a median of 4 days (IQR = 2–14 days) (Table 3). Among the 23 participants who engaged in heavy cleanup activities, 10 (43%) wore a full-face, half-face, or N-95 respirator,* half of whom reported always wearing a respirator during cleanup. Eighteen participants wore gloves during heavy cleanup, eight wore boots, and two wore goggles. Three participants reported using no PPE. Among the 27 participants who engaged in light cleanup activities, two wore a respirator, both of whom reported wearing it at all times during cleanup. Thirteen wore gloves, two used goggles, and one wore boots; seven used no PPE. Among all participants who engaged in cleanup activities, only one wore all PPE recommended for otherwise healthy persons.

Among all 103 participants, 62 (60%) reported hearing or reading about what to wear to clean up mold and floodwater. The most commonly reported information sources included television (14), word of mouth (14), and health care providers (seven). No participants reported obtaining information from social media or a website.