Reporting of False Data During Ebola Virus Disease Active Monitoring—New York City,January 1,2015-December 29,2015

Methods

Results

There were 4,726 unique persons who underwent EVD active monitoring by DOHMH during January 1 to December 29, 2015, none of whom were classified as “high risk,” and 4,075 (86.2%) of whom met eligibility criteria and were sent the survey; a total of 393 persons (9.6%) returned the survey and were included for analysis. There were no significant differences between respondents and nonrespondents in age and gender, but there were significant differences in EVD-affected country visited, country of citizenship, preferred language, and risk category (see Appendix Table).

Discussion

We report the frequency of reporting of false data during EVD active monitoring, factors related to reporting of false temperature data while being actively monitored for EVD, and the psychosocial impact of being actively monitored for EVD. The most compelling and robust finding was that more than 10% of respondents admitted to providing false data of some sort. Specifically, 14.0% of respondents admitted to having provided false temperature data during active monitoring, more than a quarter of whom reported false temperature data often or always. Others admitted to having not reported during active monitoring when they experienced EVD-like symptoms, traveled overnight outside of NYC, or went to a hospital or emergency room. It is possible that these instances of false data reporting occurred because people self-screened their risk of EVD: They believed that they did not have EVD and therefore reported false data to prevent a public health investigation. In a setting with extremely low EVD incidence, such as NYC, such self-screening will likely not threaten the overall effectiveness of active monitoring. However, increased communication to people undergoing active monitoring regarding what to expect when they report an illness might help to promote compliance with active monitoring and reduce self-screening.

Results of the multivariable analysis indicate that EVD-affected country visited, age, and rating the importance of active monitoring as low were independently associated with reporting false temperature data. It is unclear exactly why these factors may be associated with noncompliance. Regardless, these data point to an opportunity for intervention during future active monitoring operations, such as targeting communication efforts specifically toward people who have traveled to multiple EVD-affected countries and people under age 50. To promote compliance, it may also be useful to remind all people under active monitoring about the importance of active monitoring and the potential ramifications of falsifying data.

Interestingly, results of the univariate analysis indicate that the odds of reporting of false temperature data were higher among those traveling for business- or service-related reasons compared with those who traveled to visit friends or relatives. This finding is paradoxical given that some of these people who traveled for business- or service-related reasons were likely involved in some aspect of the EVD response, and one would expect them to be relatively well informed about the importance of adhering to public health recommendations; of note, this relationship lost significance after multivariable adjustment.

Another notable finding was the number of respondents who experienced workplace discrimination, with approximately 5% reporting being treated differently by someone at work, more than three-quarters of whom were told they could not come to work at all. This finding is alarming, considering that each person we monitored was told at the time of enrollment in our active monitoring program that they could continue to work even while being actively monitored (assuming they had no illness). Starting in April 2015, each person was also mailed a flyer that outlined employment rights and obligations pertaining to EVD. ¹⁴ These findings underscore that during future active monitoring operations, public health agencies could routinely inquire about workplace discrimination and help to address the situation promptly. They also highlight the dilemma for health departments of needing to educate supervisors and co-workers without infringing on the confidentiality of those under active monitoring and without inciting worry in the workplace. One potential solution is for health departments to proactively prepare informational handouts for people undergoing active monitoring to provide to their colleagues that helps reassure colleagues that the health department is working closely with them to verify that they are noninfectious.

Similarly, approximately 13% of respondents reported experiencing a behavior change among noncohabitors, corroborating results from a survey of people under EVD active monitoring in Washington, DC. ¹¹ As with workplace discrimination due to being monitored for EVD, perceived behavior changes by noncohabitors has potential implications for the mental health of people undergoing EVD monitoring. To help reassure noncohabitors, health departments could proactively prepare informational handouts that people undergoing active montioring could distribute as needed.

An unexpected finding was that more than twice as many respondents preferred to be monitored by telephone compared with the internet. Respondents also rated the cellular telephone they received at the airport on arrival to NYC as useful. This may have been influenced by respondents' overall positive ratings of their interactions with DOHMH staff and their familiarity with DOHMH EVD active monitoring procedures, given that DOHMH conducted EVD active monitoring entirely over the telephone. While some health departments have reported positive experiences using internet or short message service–based reporting, there are inherent advantages to both methods. Given respondents' preference to be monitored by telephone and the rating of the cellular telephone they received as useful, it may be beneficial for future active monitoring programs to use both options.^6,8,11,15

Overall, respondents rated their monitoring experience and interactions with DOHMH staff as positive, citing little disruption posed by active monitoring to their daily life and reporting low levels of worry about getting sick with EVD. A notable finding is the large number of respondents (39%) who cited DOHMH as helping them cope with being actively monitored, underscoring the importance of effective communication and support between those undergoing active monitoring and trained public health staff. Many respondents also rated the importance of active monitoring as high, suggesting that they heeded the messages relayed by airport-based staff and DOHMH staff. That said, some respondents also reported experiencing the negative psychological impact of being actively monitored, with more than a quarter of respondents reporting experiencing annoyance and approximately 15% reporting experiencing frustration and stress.

The higher odds of reporting of false temperature data among those indicating that they did not think it was very important to actively monitor people who recently returned from EVD-affected countries, as well as among those who reported experiencing frustration and annoyance as a result of being actively monitored, suggests that the psychosocial impact of active monitoring should be considered during future operations. Even before administering this survey, DOHMH recognized the possibility that people under active monitoring could experience stress, anxiety, or a more significant mental health emergency. Thus, we developed a protocol to refer individuals under active monitoring in need of help to a free, confidential mental health service; ultimately, we did not identify any individuals who required referral. ¹⁶ Future active monitoring programs in other jurisdictions could consider developing similar protocols to refer individuals under active monitoring to mental health services. Another possibility is to inform all individuals at the time of enrollment about the anticipated stress and anxiety of being monitored and to provide information about mental health support services.

There are several limitations inherent in our survey design. Perhaps the most significant limitation is our low response rate of 9.4%; caution should be used when interpreting our data or when making generalizations about the entire population that underwent EVD active monitoring, particularly when interpreting the findings related to psychosocial impact. Relatedly, the survey was likely prone to nonresponse bias, as those who felt compelled to complete and return the survey may have generally been more compliant with active monitoring. Moreover, the sensitive nature of the survey questions likely elicited response bias, which could mean that our results actually underrepresent frequency of reporting of false data. Also, due to data limitations, we excluded people monitored during October to December 2014, which was shortly after CDC's release of interim EVD guidance for monitoring and movement of individuals with potential exposure to EVD and which was during the height of fear in the United States, potentially biasing our results.^5,17,18 Additionally, particularly for respondents who experienced an 8-month lapse between completing active monitoring and receiving the survey, recall bias is possible. Selection bias was also likely a limitation, as some groups either were less likely to receive and complete the survey (eg, people with limited access to email or who do not speak English or French) or were not sent the survey at all (eg, people residing outside of the United States). This limitation is evident in the comparison between respondents and nonrespondents, which shows significant differences between the 2 groups in EVD-country visited, citizenship, preferred language, and risk assessment.

Active monitoring was a critical component of the domestic and global EVD response and risk management framework, helping to safeguard the public while minimizing the restriction of individual liberties: Nearly 30,000 persons were actively monitored for EVD in the United States, none of whom received an EVD diagnosis, and no cases of community transmission occurred.^3,5,17,18 The benefits of EVD active monitoring extend beyond a health department's ability to maximize public health by rapidly identifying imported cases. The centralization and interagency collaboration inherent in active monitoring allows health departments to monitor discreetly (rather than in a quarantine setting), to use trained public health physicians for clinical consultation (rather than primary care physicians evaluating patients in their office or in an urgent care setting), and to facilitate safe, efficient access, when appropriate, to healthcare facilities competent in managing EVD. It is possible that active monitoring will be needed for future infectious disease threats. As such, evaluating the factors related to reporting of false data and the psychosocial impact of active monitoring, as we did here, could potentially enhance the effectiveness of such programs in the future. While this survey provides definitive evidence confirming that people under active monitoring occasionally report false data, it remains unknown whether reporting of false data during active monitoring in low EVD-burden settings actually threatens the effectiveness of this countermeasure.