The Clinical Impact and Cost-effectiveness of Measles-Mumps-Rubella Vaccination to Prevent Measles Importations Among International Travelers From the United States

Abstract

Background

Measles importations and the subsequent spread from US travelers returning from abroad are responsible for most measles cases in the United States. Increasing measles-mumps-rubella (MMR) vaccination among departing US travelers could reduce the clinical impact and costs of measles in the United States.

Methods

We designed a decision tree to evaluate MMR vaccination at a pretravel health encounter (PHE), compared with no encounter. We derived input parameters from Global TravEpiNet data and literature. We quantified Risk_exposure to measles while traveling and the average number of US-acquired cases and contacts due to a measles importation. In sensitivity analyses, we examined the impact of destination-specific Risk_exposure, including hot spots with active measles outbreaks; the percentage of previously-unvaccinated travelers; and the percentage of travelers returning to US communities with heterogeneous MMR coverage.

Results

The no-encounter strategy projected 22 imported and 66 US-acquired measles cases, costing $14.8M per 10M travelers. The PHE strategy projected 15 imported and 35 US-acquired cases at $190.3M per 10M travelers. PHE was not cost effective for all international travelers (incremental cost-effectiveness ratio [ICER] $4.6M/measles case averted), but offered better value (ICER <$100 000/measles case averted) or was even cost saving for travelers to hot spots, especially if travelers were previously unvaccinated or returning to US communities with heterogeneous MMR coverage.

Conclusions

PHEs that improve MMR vaccination among US international travelers could reduce measles cases, but are costly. The best value is for travelers with a high likelihood of measles exposure, especially if the travelers are previously unvaccinated or will return to US communities with heterogeneous MMR coverage.

A pretravel measles-mumps-rubella (MMR) vaccination could offer good value or be cost saving for travelers with a higher risk of measles exposure, especially if they are previously unvaccinated or returning to a US community with heterogeneous MMR coverage.

Measles remains a public health challenge in the United States and is highly contagious. Clinical illness will result in 90% of unvaccinated, exposed individuals without other evidence of measles immunity [1]. When a measles case is identified in the United States, expeditious and extensive contact tracing is required, which is costly [2]. Non-specific early symptoms may cause measles to go unrecognized at an initial health care–provider visit, leading to more exposures, US-acquired cases, and contacts that require investigation [3]. Domestic outbreaks in 2011 alone are estimated to have cost local and state public health departments more than $5M and 83000 personnel hours [4]. Additional costs include hospital- or clinic-based infection control efforts and costs of clinical care or missed work for those ill with measles [5].

Although endemic transmission of measles has been eliminated from the United States since 2000 [6], new US-acquired cases continue to result from importations [7, 8]; 74% of importations in 2009–2014 were returning US residents [7]. Because US international travelers have an increased risk for measles when abroad [9], the Advisory Committee on Immunization Practices (ACIP) recommends 2 lifetime measles-mumps-rubella (MMR) vaccinations for international travelers born after 1956 and aged ≥1 year who are without other evidence of immunity, and 1 MMR vaccination for travelers aged 6 to 12 months [1, 10]. MMR vaccination reduces the potential for measles outbreaks through 2 mechanisms. First, it markedly reduces the risk that exposed individuals will become ill with measles. Second, the rare traveler who develops disease after 2 MMR vaccinations is much less likely to transmit measles, thus limiting subsequent generations of spread and reducing the number of US-acquired cases and contacts [11].

A pretravel health encounter (PHE) can reduce measles importations to the United States, if the provider assesses the traveler’s history of MMR vaccinations and vaccinates with MMR according to ACIP recommendations. However, travelers may not seek pretravel health advice, depending on their itinerary, insurance status, and personal beliefs. Even when travelers attend PHEs, clinicians do not always follow ACIP recommendations regarding pretravel MMR vaccination, most often because of traveler refusal or an erroneous provider decision [12]. Although measles is a travel-related illness, it is relatively uncommon; only 20–80 imported cases of measles are reported annually, despite more than 60M US residents traveling internationally each year [7, 13]. Thus, it is unclear whether preemptive spending to assess and vaccinate departing travelers is an optimal use of limited public health resources. We designed a decision tree model to project the clinical impact, costs, and cost-effectiveness of PHE, including MMR vaccination for eligible US international travelers, and to highlight those travelers for whom this approach offers the greatest value.

METHODS

Analytic Plan

We developed a model to simulate a cohort of US international travelers, using 2009–2014 data from Global TravEpiNet (GTEN) [14]—a consortium of 24 clinical sites throughout the United States that provide pretravel health advice and vaccination—and the Centers for Disease Control and Prevention (CDC) measles surveillance to inform input data and assumptions. We compared 2 strategies for travelers before departure: (1) no PHE and (2) PHE, in which a traveler is evaluated for baseline MMR vaccination status and vaccinated with MMR, if eligible.

We examined a cohort of 10M US international travelers, given a risk of measles exposure (Risk_exposure) based on a weighted distribution of destinations. We restricted our analysis to US travelers born after 1956, because ACIP recommendations assume a preexisting immunity among US residents born in or before 1956, when measles was widespread [1].

Model-based outcomes included the number of imported and US-acquired measles cases, costs, and incremental cost-effectiveness ratios (ICERs, net cost per measles case averted). We adopted the societal perspective and included the direct costs of clinical care, public health investigations, and indirect costs due to lost wages for travelers or caregivers. We considered ICERs of $50000 and $100000 per averted measles case as benchmarks for the value of PHE compared to no PHE. We considered the PHE strategy to be cost saving if the overall averted costs associated with improved clinical outcomes exceeded the cost of PHEs.

The study protocol (2008P001508) was reviewed by the Partners Human Research Committee and met the criteria for exempt classification.

Model Structure

We designed a decision tree (TreeAge, Williamstown MA) to simulate those US international travelers evaluated at PHE compared with those not. Travelers have a history of 0, 1, or 2 MMR vaccinations at baseline. In the no PHE branch (Figure 1, top), travelers depart without further evaluation or vaccination. During travel, they have a destination-dependent Risk_exposure; their likelihood of becoming ill with measles if exposed depends on their vaccination status. Unvaccinated travelers who become ill with measles can transmit measles to others, resulting in a greater number of contacts requiring tracing and evaluation. Travelers vaccinated once with MMR are less likely to become ill with measles if exposed, but can transmit measles if infected. Travelers vaccinated twice with MMR are even less likely to become ill if exposed, and no US-acquired cases result [11]. In the PHE branch (Figure 1, bottom), all travelers are assessed for their past history of MMR vaccinations, and 47% of MMR-eligible travelers are vaccinated [12]. The travel-related risks of measles exposure, illness, US-acquired cases, and contacts are the same in both strategies, stratified by travelers’ vaccination status.

Figure 1.

Decision tree to assess the clinical impact and cost-effectiveness of the pretravel health encounter (PHE) for measles-mumps-rubella (MMR) vaccination. Simulated travelers enter the model at the square decision node (far left), where they experience 1 of 2 strategies: no PHE or PHE. With no PHE (top), travelers depart, having had 2, 1, or 0 MMR vaccination(s) at baseline. If they attend PHE (bottom), travelers are assessed for baseline MMR vaccination status and vaccinated, if eligible. In both strategies, all travelers can be exposed to measles during travel. If exposed, travelers can fall ill with measles, or not, depending on their vaccination status. Among travelers who become ill with measles, travelers who remained unvaccinated (or are vaccinated once) will transmit the disease; travelers vaccinated twice do not. Contacts will result from any imported case of measles. Outcomes of each branch of the tree are listed to the right. Abbreviations: MMR, measles-mumps-rubella; PHE, pretravel health encounter.

Input Parameters

Demographics and Baseline Vaccination Status

Based on GTEN data (2009–2014), simulated travelers had a median age of 30 years; 56% were female (Table 1). Using data from GTEN and the National Health and Nutrition Examination Survey (2009–2010), we estimated that 7% of US travelers were unvaccinated for MMR at baseline, 9% had 1 MMR vaccination, and 84% had 2 MMR vaccinations or other evidence of immunity (Supplementary Material) [6, 12]. We excluded travelers who were not eligible for MMR vaccination (eg, pregnant).

Table 1.

Base Case Input Parameters for a Model Assessing the Clinical Impact and Cost-effectiveness of Measles-Mumps-Rubella (MMR) Vaccination During Pretravel Health Encounter (PHE)

Variable		Base Case	No. (Range)
Demographics [12]
Median age, years (IQR)		30	(22–41)
Female (%)		56
Baseline MMR vaccination status [6, 12]
2 MMR vaccinations (%)a		84
1 MMR vaccination (%)		9
0 MMR vaccinations (%)		7	(0–90)
Risk of measles exposure during travel; see Supplementary Material (exposures per 10M travelers) [15]
All international travel		237	(10-10000)
By continent
Travel to North America		19
Travel to Europe		434
Travel to Africa		681
Travel to Asia		975
Risk of measles exposure during travel; see Supplementary Material (exposures per 10M travelers) [15]
By hot spot
Travel to the Philippines, 2014		1940
Travel to India, 2011		2490
Travel to Poland, 2013		3340
Travel to Pakistan, 2013		6750
	Number of MMR vaccinations at baseline
	2	1	0
Measles infection, if exposed (%) [1, 7]	3 (0–10)	7 (3–90)	90 (0–90)
US-acquired cases, if infected (n) [3, 7]	0 (0–20)	4 (0–382)	4 (0–382)
Contacts, if infected (n) [4]	200 (0–2000)	1500 (0–12000)	1500 (0–12000)
Costs of PHE ($) [16, 17]b, c
First visit	7 (5–300)	95 (5–300)	95 (5–300)
Second visit	—	—	90 (5–300)
Costs (US$)
Per imported measles case [7, 18–20]c		13200	(11600–46800)
Per US-acquired measles case [7, 18–20]c		4800	(2500–14100)
Per contact [4, 21],c		550	(370–850)

Abbreviations: IQR, interquartile range; MMR, measles-mumps-rubella; PHE, pretravel health encounter

^aOr alternate evidence of immunity.

^bProrated for percentage of visit dedicated to measles, and includes cost of MMR vaccination and administration when given, as well as time lost from work.

^cCosts include time lost from work.

MMR Vaccination

In the PHE strategy, 47% of eligible travelers appropriately receive MMR, as per ACIP recommendations [12].

Measles Exposure During Travel

Given the number of measles importations from and the number of US travelers to each continent (2009–2014), we estimated an overall Risk_exposure, weighted by all destinations (237 exposures/10M travelers; Supplementary Material) [15, 22, personal communication with CDC]. Continent-specific Risk_exposure ranged from North America (19 exposures/10M travelers) to Asia (975 exposures/10M travelers). We also defined hot spots, or destinations with a high risk for measles exposure, as countries from which the greatest number of importations by US travelers occurred in any given year (2009-2014).

Measles Illness, If Exposed

If exposed, 90% of unvaccinated travelers without other evidence of measles immunity become ill with measles. This risk is reduced to 7% among exposed travelers with 1 MMR vaccination and 3% among travelers with 2 MMR vaccinations [7].

US-acquired Cases

We estimated that an average of 4 US-acquired cases occur per importation, based on the ratio of all importations to US-acquired cases (Supplementary Material). We assumed no US-acquired cases result from the rare returning traveler infected with measles who had been vaccinated twice, given that only 2 transmissions have ever been reported from such individuals [23, 24].

Contacts

We estimated a greater number of contacts (1500 contacts) requiring investigation when importations are due to unvaccinated or once-vaccinated travelers, because contacts need to be investigated from not only the imported case, but also any US-acquired cases [4]. Fewer contacts (200 contacts) result from travelers vaccinated twice, given that contacts result only from the importation in the absence of US-acquired cases. Given the broad range in the number of contacts per importation and/or outbreak [4, 23, 25–29], we varied these estimates widely in sensitivity analyses.

Costs

Because only 3% of PHE is devoted to measles, we prorated the cost of the clinical visit ($5) and of the missed work while attending the PHE (2 hours, $2; Supplementary Material) [16]. Those vaccinated at PHE incur additional costs for 1 or 2 MMR vaccinations ($63 each) and their administration ($25 each), depending on the number needed to achieve ACIP recommendations [17].

We estimated 1 case of imported measles to cost $13200, due to a weighted average of inpatient ($30600) and outpatient ($160) management, including the indirect costs of missing work while ill, quarantined, or caring for an ill dependent ($2000) [16, 18, 30]. We estimated a lower cost per US-acquired measles case ($4800), because fewer US-acquired cases required hospitalization when compared with imported cases (10% vs. 38%, unpublished CDC data). The average cost per contact ($550) includes the costs to state and local health departments ($298) [4], as well as missed work for the contact ($252; Supplementary Material) [21].

Sensitivity Analyses: 1-Way

We investigated the impact of uncertainty in parameter estimates by performing 1-way deterministic sensitivity analyses on clinical and cost parameters, guided by published confidence intervals and plausible ranges.

Sensitivity Analyses: 2-Way

Because Risk_exposure greatly influenced the results, we assessed its impact in combination with: (1) the percentage of unvaccinated travelers; (2) the probability that MMR vaccination is given when indicated at PHE; (3) the number of US-acquired cases; (4) the number of contacts per importation; and (5) the direct costs of PHE.

Multi-way Sensitivity Analyses: Heterogeneous Vaccination Communities

The rise of non-medical exemptions has increased the number of US communities with heterogeneous rates of childhood MMR vaccination [31]. In such heterogeneous vaccination communities, an imported measles case is more likely to result in additional US-acquired cases and contacts requiring contact tracing [32–34]. We therefore investigated the impact of travelers to different destinations and returning to such a community by varying the following parameters simultaneously: (1) Risk_exposure; (2) the number of US-acquired cases per importation; and (3) the number of contacts per importation. We also examined the impact of different percentages of unvaccinated travelers returning to a heterogeneous vaccination community by varying the following parameters simultaneously: (1) the percentage of unvaccinated travelers; (2) the number of US-acquired cases per importation; and (3) the number of contacts per importation.

Model Validation

We validated model projections to ensure consistency with the imported and US-acquired cases reported to the CDC (2009–2014; Supplementary Material).

RESULTS

Base Case

Among 10M international travelers, the no PHE strategy projected 22 measles importations and 66 US-acquired cases, at a cost of $14.8M (Table 2). Clinical outcomes improved in the PHE strategy, which projected 15 measles importations and 35 US-acquired cases per 10M travelers, averting 38 measles cases per 10M travelers, at a cost of $190.3M. This higher cost represents the costs incurred by all travelers initiating a PHE in addition to the costs of the measles cases. Among all travelers, PHE (versus no PHE) yielded an ICER of $4.6M/measles case averted.

Table 2.

Outcomes of a Model Assessing Clinical Impact and Cost-effectiveness of Measles-Mumps-Rubella (MMR) Vaccination During Pretravel Health Encounter (PHE)

	Imported Measles Cases	US-acquired Measles Cases	Averted Measles Cases	Costs (US$)	ICER (US$/ Measles Case Averted)
Base case per 10M US travelers (Riskexposure, 237 exposures/10M travelers)
No PHE	22	66		14800000
PHE	15	35	38	190300000	4611000
By destination
North America per 45M travelers (Riskexposure, 19 exposures/10M travelers)
No PHE	8	24		5300000
PHE	5	13	14	821800000	59484000
Europe per 12M travelers (Riskexposure, 434 exposures/10M travelers)
No PHE	50	145		32700000
PHE	34	77	84	238000000	2440000
Africa per 1.2M travelers (Riskexposure, 681 exposures/10M travelers)
No PHE	8	23		5100000
PHE	5	12	14	24800000	1493000
Asia per 10M travelers (Riskexposure, 975 exposures/10M travelers)
No PHE	95	278		62600000
PHE	64	147	162	222100000	991000
By hot spot
Philippines (2014) per 600000 travelers (Riskexposure, 1940 exposures/10M travelers)
No PHE	11	32		7300000
PHE	7	17	19	15000000	412000
India (2011) per 800000 travelers (Riskexposure, 2490 exposures/10M travelers)
No PHE	19	56		12500000
PHE	13	29	33	21600000	283000
Poland (2013) per 170000 travelers (Riskexposure, 3340 exposures/10M travelers)
No encounter	5	16		3600000
Pretravel encounter	4	8	9	5100000	167000
Pakistan (2013) per 99000 travelers (Riskexposure, 6750 exposures/10M travelers)
No encounter	6	18		4200000
Pretravel encounter	4	10	10	4100000	Cost-saving

Abbreviations: ICER, incremental cost-effectiveness ratio; MMR, measles-mumps-rubella; PHE, pretravel health encounter.

Destination-specific Risk of Measles Exposure

For destination-specific Risk_exposure, the ICER ranged from $59.5M/measles case averted (North America, 14 cases averted, given 45M travelers) to cost-saving (Pakistan, 10 cases averted, given 99000 travelers; Table 2). Maintaining all other base case inputs, PHE resulted in an ICER of <$100000/measles case averted for travelers to destinations with Risk_exposure ≥4180/10M travelers and was cost-saving if the Risk_exposure ≥6600 exposures/10M travelers.

Other 1-Way Sensitivity Analyses

1-way sensitivity analyses on all parameters other than Risk_exposure resulted in ICERs ≥$100000/measles case averted (Supplementary Figure 1).

Sensitivity Analyses: 2-Way

PHE offered better value or was cost-saving as Risk_exposure increased and more travelers were unvaccinated (Figure 2A). PHE offered value <$100000/measles case averted for travelers to Poland (2013) if >11% were unvaccinated and became cost-saving if travelers were going to Pakistan (2013) and >7% of travelers were unvaccinated.

Figure 2.

The pretravel health encounter (PHE) offers better value if more travelers are unvaccinated at baseline or if more eligible travelers receive measles-mumps-rubella (MMR) vaccination at PHE. The figure shows a 2-way sensitivity analyses of the value of PHE, compared with no PHE, while modifying Risk_exposure (y-axis) and the percentage of travelers unvaccinated at baseline (x-axis, A) or receiving vaccination at PHE (x-axis, B). A white X marks the base case. Horizontal lines represent the risk of exposure to measles while traveling to different continents or hot spots, from the highest (top) to lowest (bottom) risk of exposure. (A) PHE offers better value (yellow or orange) or is cost-saving (green) when Risk_exposure is high (top) or when more travelers are unvaccinated at baseline (right). (B) PHE offers better value or is cost-saving if more MMR-eligible travelers are vaccinated at PHE (right), especially for travelers whose destinations have a higher Risk_exposure (top). PHE is never preferred, regardless of the risk of measles exposure, when no MMR-eligible travelers receive the MMR vaccine during PHE (black). Abbreviations: ICER, incremental cost-effectiveness ratio, in US$ per measles case averted; MMR, measles-mumps-rubella; PHE, pretravel health encounter.

PHE provided better value for travelers with higher Risk_exposure when >47% of eligible travelers were vaccinated to achieve ACIP recommendations (Figure 2B). PHE offered value <$100000/measles case averted if at least 98% of eligible travelers to Poland were vaccinated, and remained cost-saving as long as 45% of eligible travelers to Pakistan were vaccinated. However, at a Risk_exposure ≤3330 exposures/10M travelers, PHE never resulted in an ICER <$100000/measles case averted, even when all MMR-eligible travelers were vaccinated.

The ICER for PHE was <$100000/measles case averted for travelers at a Risk_exposure as low as 262 exposures/10M travelers or 790 exposures/10M travelers, but only if ≥196 US-acquired cases or ≥11952 contacts needed investigation per importation, respectively (Supplementary Figure 2A and 2B). PHE offered improved value if the direct costs of PHE could be decreased (ie, evaluation via a phone call; Supplementary Figure 2C).

Multi-way Sensitivity Analyses: Heterogeneous Vaccination Communities

PHE could be cost-saving when each importation resulted in more US-acquired cases or extensive contact tracing, which can occur when travelers return to US communities with heterogeneous MMR coverage (Figure 3). For travelers with lower Risk_exposure itineraries (eg, Europe, Figure 3A), PHE resulted in ICERs ≥$100000/measles case averted if small or moderate-sized outbreaks resulted from an importation, such as were reported in Tennessee (Figure 3, star) or Arizona (Figure 3, square), respectively [26, 27]. PHE resulted in an ICER <$50000/measles case averted only if a large outbreak resulted from the importation, such as occurred in an Ohio Amish community (Figure 3, circle) [28]. At a higher Risk_exposure, such as in cases of travel to Africa or Asia (Figures 3B and 3C), PHE resulted in ICERs <$50000/measles case averted to prevent large outbreaks (Figure 3, triangle and circle) [28, 29]. For travelers to hot spots such as the Philippines in 2014 (Figure 3D), PHE offered an ICER <$100000/measles case averted to prevent moderate-sized outbreaks and was cost-saving to prevent large outbreaks. PHE was more likely to result in an ICER <$100000/measles case averted if more travelers were unvaccinated at baseline and returned to heterogeneous vaccination communities (Supplementary Figure 3).

Figure 3.

A pretravel health encounter (PHE) for measles-mumps-rubella (MMR) vaccination offers good value, as compared with no PHE, for travelers returning to a heterogeneous-vaccination US community, especially if returning from destinations with a high risk of measles exposure. A multi-way sensitivity analysis displays the impact of contacts (y-axis) and US-acquired cases (x-axis) due to measles importations with 0 or 1 MMR vaccination. Each panel displays the results if the travelers are returning from a different itinerary: (A) Europe (Risk_exposure, 434 exposures/10M travelers); (B) Africa (Risk_exposure, 681 exposures/10M travelers); (C) Asia (Risk_exposure, 975 exposures/10M travelers); and (D) Philippines (Risk_exposure, 1940 exposures/10M travelers). In each panel, a black X marks the base case (4 US-acquired cases and 1500 contacts per imported case from a traveler with 0 or 1 vaccination). Additional symbols denote the number of US-acquired cases and contacts reported in recent US outbreaks. Abbreviations: ICER, incremental cost-effectiveness ratio, in US$ per measles case averted; MMR, measles-mumps-rubella; PHE, pretravel health encounter.

DISCUSSION

This model-based analysis demonstrates that PHEs could greatly reduce measles cases in the United States, but such a strategy is costly when broadly applied, given the total number of US international travelers. PHEs provided better value when targeted to US travelers visiting destinations with higher risks of measles exposure, especially if these travelers were previously unvaccinated (without other evidence of measles immunity) or were returning to US communities with heterogeneous MMR vaccination coverage.

The risk of measles exposure during travel was the most influential parameter in our analysis. Unless the risk of exposure was ≥4180 exposures/10M travelers, the PHE strategy was clinically effective but never achieved value <$100000/measles case averted. Pretravel MMR vaccination should be prioritized for travelers to destinations with higher exposure risks, such as countries with endemic disease or active outbreaks, as reported by CDC travel notices [35]. The strategy would provide even better value among such travelers if the MMR vaccination of eligible travelers at the encounter is improved beyond the 47% currently reported [12] or if the costs of evaluation could be further reduced (ie, evaluation via a phone call).

PHE offers substantial clinical impact, even for travelers to destinations with lower risks of exposure. For example, Europe was the source of more than a quarter of all US measles importations in 2009–2014 [7]. However, an estimated 12M US travelers visited Europe during that time, so the per-traveler risk of measles exposure was lower than for other destinations. This model-based analysis, therefore, projected that a PHE for travelers to Europe would be clinically effective (averting 84 cases per 12M travelers), but at an ICER of $2.4M/measles case averted, because so many travelers would need to be assessed and vaccinated.

PHEs offered the greatest clinical benefit for travelers who were unvaccinated at baseline: they could offer good value and even be cost-saving in this subpopulation, if the risk of exposure was sufficiently high. Unvaccinated travelers may be the most challenging group to convince of the benefits of MMR vaccination, given religious or personal belief justifications for MMR refusal. However, patients who previously refused a vaccination are known to accept vaccination more readily in the face of exposure or outbreak [36]. These model-based results, therefore, support ongoing efforts to improve the education of travelers and providers regarding the risks of measles during travel and the role and safety of MMR vaccination. Such discussions may be well suited to a primary care visit with an established therapeutic relationship, which might be particularly important in addressing concerns of unvaccinated travelers [36–38].

If travelers returned to heterogeneous vaccination communities, the impact of an imported measles case was magnified dramatically, given that more US-acquired cases and contacts would be expected to occur [8, 28, 29]. PHE offered excellent value to prevent large measles outbreaks [28, 29], and even smaller outbreaks [26], if the risk of measles exposure was high. For travelers to specific hot spots in certain years, PHE could offer good value to prevent even relatively small outbreaks [27]. Improving MMR vaccination among travelers returning to US communities with heterogeneous vaccination coverage should be prioritized, with focused interventions that highlight the benefits of MMR vaccination, such as the intervention that occurred after the 2015 Disneyland outbreak [36].

Clinically-effective healthcare interventions that protect the public are often not cost-effective, yet can still be important for public health. Given the extreme infectivity of measles and its airborne transmission [39], travelers can be exposed to measles in airports or through interactions with travelers from endemic countries [40]. Therefore, the current ACIP recommendation for all US international travelers to have presumptive evidence of immunity against measles is a conservative public health approach. However, such an approach is likely not cost effective, especially for itineraries to North or South America, which included most US international travelers (63%) and the fewest measles importations (4%) in 2009–14.

This analysis includes important assumptions and limitations. The number of measles importations reported to the CDC is likely an underestimate. Thus, the risk of measles exposure may be higher than estimated, which would improve the value of PHE. Our model structure assumed that unvaccinated travelers are contagious with measles after their return to the United States, given the incubation period (7–21 days) [39], duration of contagiousness (9 days) [39], and mean duration of US residents’ international travel (18 days) [41]; however, we validated model outcomes to reported measles importations. The analysis did not account for the additional benefits of increasing herd immunity by also improving travelers’ immunity to mumps and rubella, which would further increase the strategy’s clinical benefit and cost effectiveness, given recent outbreaks of mumps [42] and the impact of congenital rubella infection [43]. PHE includes additional health benefits and costs not addressed in this measles-focused analysis [14] and provides the best value for recurrent international travelers, since measles immunity is lifelong. Lastly, travelers who attend PHE could differ from those who do not in terms of planned itinerary, health insurance, and willingness to accept recommendations [12].

PHEs that include MMR vaccinations for those eligible are costly but could reduce US measles cases and are a worthwhile public health intervention. In the absence of resources to cover all international travelers, efforts should be focused where they will be most clinically effective and provide the best value: travelers to higher-risk destinations, especially those travelers who are previously unvaccinated or are returning to US communities with heterogeneous MMR vaccine coverage.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Acknowledgments. Members of the Global TravEpiNet Consortium include George M. Abraham, Saint Vincent Hospital (Worcester, MA); Salvador Alvarez, Mayo Clinic (Jacksonville, FL); Vernon Ansdell and Johnnie A. Yates, Travel Medicine Clinic, Kaiser Permanente (Honolulu, HI); Elisha H. Atkins, Chelsea HealthCare Center (Chelsea, MA); Holly K. Birich and Dagmar Vitek, Salt Lake Valley Health Department (Salt Lake, UT); John Cahill, Travel and Immunization Center, St. Luke’s-Roosevelt (New York, NY); Lin Chen, Mount Auburn Hospital (Cambridge, MA); Bradley A. Connor, New York Center for Travel and Tropical Medicine, Cornell University (New York, NY); Roberta Dismukes, Jessica Fairley, Phyllis Kozarsky, and Henry Wu, Emory TravelWell, Emory University (Atlanta, GA); Ronke Dosunmu, JourneyHealth (Maywood, NJ); Jeffrey A. Goad and Edith Mirzaian, International Travel Medicine Clinic, University of Southern California (Los Angeles, CA); Nelson Iván Agudelo Higuita, University of Oklahoma Health Sciences Center (Oklahoma City, OK;) Karl Hess, Hendricks Pharmacy International Travel Clinic (Claremont, CA); Noreen A. Hynes, John Hopkins Travel and Tropical Medicine, Division of Infectious Diseases, John Hopkins School of Medicine (Baltimore, MD); Frederique Jacquerioz and Susan McLellan, Tulane University (New Orleans, LA); Jenn Katsolis, Jacksonville Travel Clinic-St. Vincents (Jacksonville, FL); Paul Kelly, Bronx Lebanon Medical Center (New York, NY); Mark Knouse, Keystone Travel Medicine, Lehigh Valley Health Network (Allentown, PA); Jennifer Lee, Northwestern Medical Group-Travel Medicine, Northwestern Memorial Hospital (Chicago, IL); Daniel Leung, Brian Kendall, and DeVon Hale, International Travel Clinic, University of Utah (Salt Lake City, UT); Alawode Oladele and Hanna Demeke, DeKalb County Board of Health Travel Services, DeKalb North and Central-T.O. Vinson Centers (Decatur, GA); Alawode Oladele and Althea Otuata, DeKalb County Board of Health Travel Services, DeKalb East (Decatur, GA); Roger Pasinski and Amy E. Wheeler, Revere HealthCare Center (Revere, MA); Adrienne Showler, Laura Coster, and Jessica Rosen, Infectious Diseases and Travel Medicine, Georgetown University (Washington, DC); Brian S. Schwartz, Travel Medicine and Immunization Clinic, University of California (San Francisco, CA); William Stauffer and Patricia Walker, HealthPartners Travel Medicine Clinics (St. Paul, MN); and Joseph Vinetz, Travel Clinic, Division of Infectious Diseases, Department of Medicine, University of California-San Diego School of Medicine (La Jolla, CA). The authors thank Audrey Bangs for assistance with manuscript preparation.

Disclaimer. The content is solely the responsibility of the authors, and the study’s findings and conclusions do not necessarily represent the official positions of the Centers for Disease Control and Prevention (CDC) or the National Institutes of Health (NIH).

Financial support. This work was supported by the US CDC (grant numbers U19CI000514, U01CK000175, and U01CK000490); the NIH (grant numbers R01AI42006 and K01HL123349 to E. P. H.), and by the Massachusetts General Hospital (the Steve and Deborah Gorlin Research Scholars Award to R. P. W.).

Potential conflicts of interest. E. T. R. reports grants from the CDC, during the conduct of the study. N. F. F. reports grants from the NIH, during the conduct of the study. E. P. H. reports grants from the NIH and CDC, during the conduct of the study. R.P.W. reports grants from the NIH and was a recipient of the Massachusetts General Hospital Steve and Deborah Gorlin Research Scholar Award, during the conduct of the study. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.