CDC Report as Required by the 2017 Control of Communicable Diseases Final Rule

Background

The Department of Health and Human Services (HHS), Centers for Disease Control and Prevention (CDC), published the final rule for the Control of Communicable Diseases on January 19, 2017, which included, among other provisions, amendments to the foreign quarantine regulations for the control of communicable diseases. The rule became effective on March 21, 2017. See 82 FR 6890.

CDC regulations at 42 CFR 71.4 (airlines) and 42 CFR 71.5 (vessels) relate to the transmission of passenger, crew, and flight/voyage information to CDC for public health purposes. Under these regulations, the operator of any airline or vessel arriving into the United States must make specified passenger and crew contact information available to CDC, to the extent that such information is available and already maintained by the operator, within 24 hours of a CDC request. This request is made only after CDC has determined the presence of a confirmed or suspected case of a communicable disease on board an aircraft or ship. For more information about CDC’s contact investigations for communicable diseases on aircraft, see Protecting Travelers’ Health from Airport to Community: Investigating Contagious Diseases on Flights.

42 CFR 71.4 (airlines) and 42 CFR 71.5 (vessels) both contain subsections that state:

No later than February 21, 2019, the Secretary or Director will publish and seek comment on a report evaluating the burden of this section on affected entities and duplication of activities in relation to mandatory passenger data submissions to [U.S. Department of Homeland Security, Customs and Border Patrol] DHS/CBP. The report will specifically recommend actions that streamline and facilitate use and transmission of any duplicate information collected.

This report evaluates the potential duplicative burdens that these regulatory provisions may have created on the airline and ship industries since they entered into effect on March 21, 2017. It also makes additional recommendations based on these findings and solicits public comment. CDC accepted public comment from February 12, 2019 through March 14, 2019.

Partner Outreach

On February 26, 2018, air and maritime experts in CDC’s Division of Global Migration and Quarantine (DMGQ) met with CDC air and maritime analysts to discuss whether any changes to data transmissions from airlines or vessels had been observed since these regulations went into effect. These subject matter experts said no changes had been noted or reported to CDC since the final rule became effective. These experts also provided a list of external partners who would be able to best inform the evaluation of additional burdens the final rule may have caused. On April 24, 2018, CDC contacted the following seven federal and non-government partners whose air or maritime operations may have been affected by the updated provisions regarding the mandatory submission of passenger data upon request.

• Airlines for America (A4A)¹
• Cruise Lines International Association (CLIA)
• Delta Air Lines, Inc.
• International Air Transport Association (IATA)
• International Civil Aviation Organization (ICAO)
• U.S. Customs and Border Protection (CBP), National Targeting Center (NTC)
• U.S. Coast Guard (USCG)

The correspondence sent to these partners included the following questions:

1. Since publication of the communicable diseases final rule on January 19, 2017, with effective date March 21, 2017, have you made any changes to your procedures in response to the publication of data collection requirements under 42 CFR 71.4 (71.5)?
2. Has the publication of the data collection requirements under 42 CFR 71.4 and 71.5 caused your agency or organization additional burden? If so, please describe any additional burden: Administrative? Operational?
   • What types of staff are involved, and approximately how much time per staff type is required (e.g., legal one hour, data management two hours)?
3. Did the publication of the requirements specified in 42 CFR 71.4 (71.5) result in duplication of activities between your agency or organization and CDC?
   • If so, please explain the new duplicative activities:
     • What types of staff are involved, and approximately how much time per staff type is required (e.g., legal one hour, data management two hours)?
4.  If you responded in the affirmative to questions 1-3, please let us know of any recommendations to minimize duplication of activities or other burdens that have resulted [from] the publication of 42 CFR 71.4 and/or 42 CFR 71.5.
   • If these recommendations were put in place, what do you expect would be the change in time spent by different types of staff (e.g., legal consultation one hour, data management two hours)?

Partner organization representatives were informed that their participation in the evaluation was strictly voluntary and would not impact the organization’s relationship with CDC. A reminder correspondence was sent on April 30, 2018.

Of the seven partners contacted, CDC received feedback from five. Four of these partners responded that their organization or agency had not experienced any increased burden or duplication of activities in response to the publication of data collection requirements under 42 CFR 71.4 or 71.5. Since no duplication was reported, these partners did not provide any recommendations to minimize duplication.

While the remaining partner also reported no additional burden beyond usual work duties associated with 42 CFR 71.4 and 71.5, the organization made the following statement: “Although we have not observed specific duplication of activities we do recommend that the CDC utilize existing data feeds provided to CBP via the Advance Passenger Information System (APIS), rather than requiring establishment of a separate channel of reporting. This would not necessarily result in a net decrease in work for our staff, but it would help lessen data privacy concerns for the airline and its customers. We have not observed a meaningful increase in workload, and as such do not have any recommendations for minimization of duplicative activities.”

CDC does make use of information from CBP’s National Targeting Center (NTC) via APIS for the purposes of public health contact investigations, to the extent CDC experts believe that NTC’s databases are the most up-to-date and accurate source of relevant data. To facilitate this process, CDC has analysts co-located at NTC who conduct data searches to supplement passenger contact information provided by airlines. However, for the following reasons, CDC must still obtain certain data first from airlines:

• Only information directly from airlines can produce a targeted partial manifest in the rows or seats most at risk of exposure from an infectious traveler. Data from CBP’s NTC can quickly produce a manifest of the entire aircraft, which is useful in only a minority of events.
• Information directly from airlines is essential to quickly identify infants in arms (information not contained in APIS) and their co-travelers. This information is critically important for certain infections for which infants may be at greater risk of serious disease or less likely to be vaccinated.
• Information directly from airlines is essential in quickly determining whether an individual remained in their assigned seat from departure to landing.
• In some cases, reconciliation of data between airlines and CBP is not complete when an individual deplanes during a layover and does not re-board.
• Finally, information from CBP does not include the configuration of an aircraft cabin (such as layout of seats and bulkheads), which is critically important when determining which passengers may have been exposed to a communicable disease.

To summarize, contacting an airline first is of the utmost importance in establishing a basic set of information about certain travelers and aircraft configuration, although there are instances when even airlines are unable to confirm travelers’ seating locations throughout flights. In keeping with current practice, to the extent CDC can obtain information from CBP, CDC will minimize additional requests to airlines.

In summary, no increased burden or duplication of effort was identified as a consequence of the 2017 final rule.

Analysis of CDC Data

Air contact investigations

Air contact investigations are usually initiated after travel, when infected travelers might be identified by a doctor or local public health department and reported to CDC. Potential contacts typically include passengers in nearby rows or seats and crew that worked in the section of the airplane where the ill traveler sat. While airlines usually assess their crew, CDC works with health departments to notify travelers about their possible exposures. After receiving input from partners that the regulations created no additional burden, CDC analyzed the timeliness and completeness of data received from airlines from before to after these regulations went into effect. In addition, CDC analyzed whether changes in timeliness and completeness of data from airlines affected the quantity of data sent to health departments after additional data searches at NTC by CDC. This analysis focused on 1) the time between manifest request to airlines and the receipt of data (timeliness), 2) the fraction of each requested data element (first name, last name, U.S. address, phone1, phone2, email address, seat #) for which data were provided by airlines (completeness), and 3) differences in the amount of data provided to health departments after CDC engaged additional resources at the NTC. The third analysis may indirectly assess the quality of data provided by airlines. The underlying assumption was that if airlines provide better data to CDC, then CDC, using supplementary data from CBP, may be able to transmit more contact information to health departments. It should be noted that the quality of data available to airlines depends on their customers providing accurate data when purchasing tickets. Since there is not a mechanism in place to ensure that travelers provide accurate contact information to airlines, airlines may not have access to accurate contact data. Airlines might also not have access to passenger contact information if the tickets were bought through third-party vendors.

This analysis focused on data provided for international flights on which a communicable disease of public health concern was reported. To assess timeliness and completeness, CDC extracted a convenience sample of contact investigation data from CDC’s Quarantine Activity Reporting System for 51 flights from the period preceding publication of the final rule (flights occurred between June 1 and December 8, 2016) and 48 to 50 flights after the regulations went into effect (flights occurred between June 13, 2017, and January 12, 2018).

The timeliness of data transmissions appeared to improve slightly after the regulations went into effect (Table 1). Requests were divided by urgency with tuberculosis investigations typically classified as non-urgent because tuberculosis cases in travelers are often diagnosed weeks to months after the travel occurred and it takes weeks to years before a person infected with tuberculosis would begin to experience symptoms. In comparison, the time to develop symptoms after exposure (incubation period) is typically much shorter for other diseases (e.g., measles, meningococcal disease [illness caused by Neisseria meningitidis bacteria], pertussis [whooping cough], and rubella [German measles]), and less time is available to provide preventive measures. Prior to the regulations going into effect, urgent requests typically specified 24 hours for airlines to return data to CDC.  In comparison, prior to the regulations, most non-urgent requests specified either 48 hours (if requests were made during the week) or 72 hours (for requests made over the weekend). However, after the regulations went into effect, all urgent and non-urgent requests specified 24 hours. For the 51 flights that occurred before the regulations went into effect, 44 investigations were non-urgent. Among the 50 flight investigations included in the post-effective-date analysis, 39 were non-urgent.

Before the regulations went into effect, airlines returned contact data for urgent requests within 24 hours for 57% of manifest requests (4 out of 7). In addition, contact data were provided after 6 days for 29% of the urgent requests (2 out of 7). In comparison, after the regulations went into effect, contact data were provided within 24 hours for 64% of the CDC requests (7 out of 11), and contact data were provided within 3 days for 100% of urgent requests.

Table 1. Time from CDC request to receipt of data from airlines before and after March 21, 2017, the effective date for 42 CFR 71.4 (airlines) and 42 CFR 71.5 (vessels)

Time from CDC request to receipt of data from airlines before and after March 21, 2017, the effective date for 42 CFR 71.4 (airlines) and 42 CFR 71.5 (vessels)

	Before the effective date (51 flights between June 1 through December 8, 2016)		After the effective date (50 flights between June 13, 2017, and January 12, 2018)
Time between request and receipt of data from airlines	Urgent requests (e.g., measles, meningococcal disease) a	Non-urgent requests (e.g., tuberculosis)	Urgent requests (e.g., measles, meningococcal disease)	Non-urgent requests (e.g., tuberculosis)
<24 hours	4 (57%)	5 (11%)	7 (64%)	9 (23%)
>=24 hrs to <72 hrs	1 (14%)	18 (41%)	4 (36%)	10 (26%)
>72 hrs to <144 hrs	0%	10 (23%)	0%	9 (23%)
>144 hrs	2 (29%)	11 (25%)	0%	11 (28%)

Note: There are too few observations to conduct statistical tests to identify differences between before and after the regulations went into effect.

^aCDC typically required that airlines provide data for urgent requests within 24 hours.

The completeness of data received was assessed based on percentage of contacts for whom first name, last name, seat number, one phone number, two or more phone numbers, email address, and complete or partial address were provided. The email and phone number fields are fairly straightforward, as information is either present or missing for each traveler. Address information is more challenging because there are a number of fields that may be partially complete. This analysis included three address categories: category 1: no address information; category 2: any address information (e.g., street address or city/state or zip code or foreign country); and category 3: complete address information (e.g., street address and city/state or zip code). In addition to traveler-specific information, airlines may include emergency phone numbers or addresses. The assumption was that this emergency contact information could be used to reach a family member or friend of the traveler. These data were analyzed separately to identify the proportion of contacts for whom traveler-specific contact information was not provided but emergency contact information was provided.

Comparing data provided before and after the regulations went into effect, less information was provided for U.S. address categories, email address, and second phone number after the regulations went into effect. More data were provided for the first phone number and an emergency contact address. There was no significant difference for other categories (Table 2).

Table 2. Completeness of airline data provided to CDC before and after March 21, 2017, the effective date for 42 CFR 71.4 (airlines)

Completeness of airline data provided to CDC before and after March 21, 2017, the effective date for 42 CFR 71.4 (airlines)

Airline manifest traveler data category	Before the effective date (51 flights from June 1 through December 8, 2016, n = 1,571)	After the effective date (48 flights from June 13, 2017, through January 12, 2018, n = 1,433)
First name	99.9%	99.9%
Last name	99.9%	99.0%
Seat number	99.7%	99.9%
U.S. address category 1 (any address information) a	28.5%*	21.1%
U.S. address category 2 (complete address information) b	20.0%*	15.9%
Emergency contact address information (if traveler information was missing)	1.7%	5.6%*
Email address	34.7%*	24.9%
Single phone number	37.3%	41.3%*
Two or more phone numbers	10.6%*	6.6%
Emergency contact phone number (if traveler phone numbers were missing)	0.1%	0%
Any information besides seat number	54.3%	49.7%

Note: * indicates that differences were significant at the 95% level based on both Fisher’s exact test for pairwise comparisons and logit models that controlled for urgency of requests and for foreign vs. domestic carriers.

^aAny address information (i.e., street address OR [city/state or zip code] OR foreign country)

^bComplete address information, (i.e., street address AND [city/state or zip code])

CDC also assessed whether more complete data were provided by CDC to health departments after the regulations went into effect by reviewing the amount of data sent by CDC to health departments after the data were supplemented by information obtained from NTC. In general, the quantity of data provided by CDC to health departments increased after the regulations went into effect (Table 3). The number of travelers with full address information increased from 84.5% of 1,462 travelers before the regulations entered into effect to 95.1% of 1,375 passenger contacts after the regulations went into effect. In addition, two or more phone numbers were provided for more travelers after the regulations went into effect (53.8% vs. 43.4%). However, no statistical differences were found among the number of contacts for whom an email address was provided or for whom at least one phone number was provided. In total, CDC was able to provide at least one piece of contact information for 99.9% of travelers both before and after the regulations went into effect.

Table 3. Contact data sent from CDC to health departments before and after March 21, 2017, the effective date for 42 CFR 71.4 (airlines)

Contact data sent from CDC to health departments before and after March 21, 2017, the effective date for 42 CFR 71.4 (airlines)

Traveler data category	Before the effective date (51 flights from June 1 through December 8, 2016, n = 1,462)	After the effective date (48 flights from June 13, 2017, through January 12, 2018, n = 1,375)
Seat number	99.7%	99.8%
U.S. address category 1 (any address information) a	84.5%	95.1%*
U.S. address category 2 (complete address information) b	93.8%	98.8%*
Emergency contact address information (if traveler information was missing)	0%	0%
Email address	78.9%	79.9%
Single phone number	92.1%	91.1%
Two or more phone numbers	43.4%	53.8%*
Phone or emergency or contact phone if traveler phone numbers were missing	1.8%	5.6%*
Any contact data	99.9%	99.9%

Note: * indicates that differences were significant at the 95% level based on both Fisher’s exact test for pairwise comparisons and logit models that controlled for urgency of requests and for foreign vs. domestic carriers.

^aAny address information ( street address OR [city/state or zip code] OR foreign country)

^bComplete address information (street address AND city/state or zip code)

Maritime contact investigations

In contrast to air contact investigations, most urgent maritime contact investigations, such as for measles, are undertaken before travelers disembark from vessels. In such instances, travelers may be informed of their exposures while they are still on vessels. In contrast, the process for air travelers is different because CDC must work with airlines to collect contact information post-travel. On occasion, cruise ship passengers might be diagnosed with, for example, tuberculosis after travel.  In this example, their contacts would include cabin-mates, dining mates, traveling companions, friends, intimate partners, or any crew members (such as waiters, cabin stewards, day care personnel) with whom they had repeated or prolonged exposure (i.e., interacted with daily or for extended periods of time). In comparison, as noted above, almost all infected air travelers are identified after travel and their contacts typically include passengers in nearby rows or seats. On average, CDC conducted about 12.6 maritime contact investigations per year from 2010 through 2014; however, only 5.4 contact investigations per year were conducted after a traveler had disembarked from the vessel. An average of 55 passenger contacts per year were identified for maritime contact investigations between 2010 and 2014 (including passengers identified during contact investigations before the ill traveler disembarked)². In contrast, when vessel crew members are identified for contact investigations, the crew are usually still on the vessel and easily located. On occasion, CDC may conduct a contact investigation for measles diagnosed after travel in which passenger contact information is sought; however, such events are so infrequent that data is very limited. Therefore, because of insufficient data, CDC did not attempt a maritime contact data analysis. In addition, CDC did not receive any feedback from vessel operators about changes in procedures or suggestions for improvement after the regulations went into effect.

Limitations

The analysis of air traveler contact data was limited by the availability of data and evolving processes to search for additional contact data at NTC. CDC did not have sufficient data to evaluate the timeliness of data receipt from airlines (Table 1) since most of the requests were not considered urgent. The analysis of completeness (Table 2) only included a subset of flights directly before and after the regulations went into effect. The numbers of investigations for each airline and for each disease varied across the two subsets. Such differences in the composition of requests may limit the ability to quantify differences directly attributable to the regulations. For example, Carrier A may always provide more contact data than Carrier B. In the before subset, there may be 5 flights with illnesses from Carrier A and 2 flights with illnesses from Carrier B. In the after subset, there may instead be 2 flights from Carrier A and 6 flights from Carrier B. In the statistical analysis, CDC attempted to control for U.S vs. non-U.S. carriers and for the urgency of requests; however, CDC was unable to control for all of the compositional differences between the before and after subsets to assess the impact of the regulation on completeness of data provided by airlines. The analysis of data sent to health departments (Table 3) may have been affected by changes in data availability at NTC or in CDC or CBP procedures at NTC that were unrelated to the regulations. As a result, the analysis of data sent by CDC to health departments should be considered as an overall assessment of the impact of the regulations plus ongoing improvements to data searching activities at NTC. Thus, the results shown in Table 3 may only be partially attributed to the regulations; some of the improvement in the amount of data sent by CDC to health departments may also have been due to improvements in NTC data searching capacity.

Discussion

Upon review of input received from federal and non-governmental partners, as well as results from the analysis of CDC data, there is some evidence that timeliness has improved since these regulations went into effect; however, completeness of contact data provided by airlines generally has not changed.  CDC found improvements for some data elements and that less data were provided for other data elements. Some airlines may be providing less complete data to improve the timeliness of data submission. However, at the same time, CDC has slightly increased the amount of data (specifically address information and second phone numbers) sent to health departments after supplementing data received from airlines with additional data obtained from CBP’s NTC. CDC cannot assess whether the larger fraction of contacts with address information or more than one phone number has improved health departments’ abilities to locate exposed travelers. In addition, CDC is unable to evaluate directly whether changes in timeliness or completeness of data provided by airlines led to the increase in data provided to health departments after supplemental searches for contact information at NTC. Finally, querying partners verified that publication of these regulations has not changed procedures or operations nor added to the burden of transmitting data to CDC for public health purposes. The analyses presented here focused on data provided by airlines because CDC does not have sufficient data for a comparable analysis of data from maritime contact investigations conducted after travel. In addition, CDC did not receive any feedback from vessel operators.

Recommendations

Because CDC did not find any evidence or receive information from partners that these regulations resulted in any additional duplication of efforts beyond that needed for a timely public health response, our recommendations are as follows:

1. CDC should continue to evaluate data collection requirements routinely to ensure that burden to respondents is limited to that needed to conduct CDC’s public health mission. To ensure this evaluation, CDC’s information collection approval for international data (OMB Control Number 0920-1180) must be renewed every three years and include requirements for soliciting public comment.
2. CDC should continue to work with partners, both federal and private sector, to improve the data collection process and minimize duplication of effort. CDC is in routine contact with aviation and maritime partners for the purposes of preventing the spread of communicable disease.  If problems or duplicative processes are brought to CDC’s attention, CDC should consider reasonable approaches to remedy the problem or reduce duplication while ensuring public health protections remain in place.
3. CDC should continue to ensure its data collection requirements reflect the most up-to-date technology and enhance overall efficiency so response times remain effective and CDC can facilitate timely public health action.

References

1. A4A shared CDC’s solicitation for feedback with all of its members: Alaska Airlines, American Airlines, Atlas Air, FedEx, Hawaiian Airlines, JetBlue, Southwest Airlines, United Airlines, and United Parcel Service.
2. Caroline E. Stamatakis, Marion E. Rice, Faith M. Washburn, Kristopher J. Krohn, Millicent Bannerman, Joanna J. Regan. (2017) Maritime illness and death reporting and public health response, United States, 2010–2014. Travel Medicine and Infectious Disease 19: 16–21.