Ongoing deployments since 2004 have affected the population dynamics at military installations and military treatment facilities (MTFs). When operational Army units such as infantry brigades deploy, active-duty health care providers assigned to the units go with them, and so do some active-duty providers who are assigned to work full-time at MTFs. So when large Army units deploy and leave the installations at which they train, the number of providers available to provide care for soldiers and other beneficiaries at the installation decreases, as does the number of beneficiaries seeking care, through the large-scale departure of soldiers deploying with the units. Under the Army Force Generation (ARFORGEN) cycle, units and large numbers of soldiers deploy and return home to installations on a predictable schedule, and during the affected time periods, the population of the installation shifts (e.g., resulting in fewer soldiers present or a changing mix of soldiers).
Army officials were concerned about the possible effects of variations (driven by deployments) on the demand for and availability of health care. In particular, the Vice Chief of Staff of the Army (VCSA) wanted to know whether Army deployments were having unintended and unknown effects on the well-being of soldiers and their immediate families. The VCSA was concerned that ebbs and flows in the ability of Army military treatment facilities to provide medical care might affect the system's ability to respond to changes in family needs as soldiers deploy, redeploy, and return home.
In March 2009, the VCSA asked RAND Arroyo Center to investigate the “collateral effects” of ARFORGEN on soldiers' and, especially, families' ability to receive health care. Aware of the deployment cycle changes in the number of available health care providers and beneficiaries seeking care, we were also cognizant of recent military health research showing that the family members of service members who deployed utilized health care in different ways (Eide et al., 2010; McNulty, 2003; Gorman et al., 2010). We designed a study to focus on two main questions:
- How does the deployment cycle affect capacity and beneficiary utilization at Army MTFs?
- How does the deployment cycle affect family health care utilization?
To answer the first question, we performed an aggregate-level analysis of deployment cycle effects on installations. This analysis included 14 installations, which accounted for 80 percent of the soldier deployments in the time period of our analysis, 2004–2009. To answer the second question, we performed a longitudinal analysis of Army families to identify how individual family utilization changes when members of the family deploy (while controlling for other factors). To perform this analysis, we assembled longitudinal records for Army family members, linking all TRICARE-eligible beneficiaries to the sponsoring active component soldier and including available demographic data for family members (age, gender, ethnicity). This analysis included the majority of the overseas contingency operations spanning 2004–2009, and the majority of active component family members who are enrolled in TRICARE Prime.
MTF-Level Analysis of Deployment Cycle Effects
We performed an analysis of 14 installations in the United States that deploy brigade combat teams (BCTs), the fundamental unit of Army deployments, and host Army hospitals. These installations are not typical military installations—they are the specific Army forts that deploy the vast majority of forces, and are referred to as force projection platforms. Over the course of this analysis, these installations generated 80 percent of deploying soldiers.* This analysis investigated changes in: beneficiary population at installations, capacity at installation MTFs, utilization at the MTF from civilian providers, and MTF provider workload.
We hypothesized that if large numbers of soldiers deployed concurrently, the dramatically diminished population of beneficiaries remaining nearby to the installations would utilize less care, in aggregate. We also expected that the total number of health care providers available at installations would decrease when Army units deployed, in that MTFs would not bring in military, civilian, or contractor providers in sufficient quantity to offset the loss. However, we did not know which change would be larger proportionally, the decrease in providers or in beneficiaries.
We hypothesized that access to care would change when the number of providers available and the number of beneficiaries competing for appointments changed, and that families of deploying soldiers might change the way they sought care. However, we did not know whether access to care would increase or decrease, and how families would seek care differently. So we could not project what changes related to access and care-seeking behavior we would observe in utilization by the beneficiaries (dependents and nondeploying soldiers) remaining near installations when Army units deployed.
Effects of Soldier Deployment on Beneficiary Population and Enrollment at the MTF
At the 14 installations in our analysis, we quantified the changes in beneficiary population sizes for the following groups: soldiers enrolled at the MTFs, who may deploy; soldier family members enrolled for care with the MTFs and with civilian network providers, who may leave the area when soldiers deploy; and, to a limited extent, retirees and their dependents who enrolled at MTFs for care, whose needs we do not expect to change related to the deployment cycle. The size of the beneficiary population is a main driver of the overall demand for care, compounded with the rate at which beneficiaries utilize care.
Soldiers account for approximately 40 percent of MTF enrollees, which is equivalent to soldier family members, who account for approximately 40 percent of MTF enrollees. Retirees and their dependents make up the balance, or approximately 20 percent of enrollees at the MTF. We present these data as a basic context to understand the extent to which changes in utilization by each group will affect overall utilization.**
Nearly all soldiers who enroll with TRICARE enroll at the MTF for care. We study changes in the enrolled soldier population and changes in the health care this population utilizes. However, we also observe that nonenrolled soldiers account for a sizable amount of care at MTFs. From discussions with MTF staff, we understand that soldiers who are temporarily assigned to an installation for training are not enrolled at the MTF. We expect that these trainees generate many of the visits by nonenrolled soldiers, but we do not study this population. Whether these visits are generated by trainees or other nonenrolled soldiers does not affect our analysis. Although we cannot quantify the nonenrolled soldier population that relies upon the MTF for care in our analysis data, we do include MTF visits by these soldiers in the analysis, as changes in these visits may magnify or more likely mitigate effects from deploying soldiers.
Over the course of the analysis, the population of beneficiaries grew, and the portion of soldiers enrolled at the MTFs grew. While the population of soldiers enrolled at MTFs in our analysis grew from 229,000 to 314,000,*** the population of family members enrolled at MTFs increased slightly less, from 292,000 to 319,000. During this time there was a large increase in family members enrolled at network providers at the installations in our analysis, an increase from 49,000 to 95,000.
Soldier deployments had major effects on the population in-garrison at the installations in our analysis. Most installations saw as many as 60 percent of the assigned soldiers deploy concurrently, emptying out the installation. We attempted to discern the extent to which soldier families left installations when units deployed but could not do so conclusively. We observe sizable changes in the population of family members living near and enrolled at MTFs, but some changes appear to be broad patterns of growth or reduction, and it is difficult to observe a clear relationship between population changes and deployments. Additionally, we note that families may not reliably update addresses and TRICARE enrollment if they move from the area temporarily.
Effects of Deployment on Health Care Provider Availability
To understand the effects of deployments on installations, we considered how deployments affect the number of providers at MTFs, the quantity of provider full-time equivalent (FTE) hours available for patient care, and provider workload.
Over the period studied (2004–2009), there was a change in the mix of providers at MTFs that was steady and not specifically related to the deployment cycle. In 2004, nearly three-fourths of the FTEs at the MTFs in the analysis were active-duty, with the largest represented group being active-duty physicians. By 2009, civilian and contractor providers accounted for nearly half of the FTEs at the MTFs, and the fastest-growing group consisted of civilian and contractor providers other than physicians, particularly civilian mental health and primary care professionals.
Providers Assigned to MTFs Typically Do Not Deploy in Large Numbers When Installation Soldiers Deploy
We observed a strong relationship between soldier deployments and deployment of health care providers assigned to the operational units (table of organization and equipment, or TOE, providers), but a weak relationship between soldier deployments and deployment of the health care providers assigned to the MTFs (table of distribution and allowances, or TDA, providers). In the 14 installations in our analysis, TDA providers assigned to MTFs did not typically deploy in significant numbers when soldiers from the same installation deployed. In contrast, TOE providers, who are generally battalion surgeons and physician assistants (PAs), deployed nearly in parallel with soldiers from the same installations. A large unit deploying from an installation may cause a decrease in the at-home soldier population by 40 percent.**** Using that as an example, when 40 percent of the soldiers at an installation deploy, only about 2 percent of providers assigned to the MTF deploy, while 32 percent of providers assigned to the installation's operational units deploy.
These effects varied across installations. “Purer” deployment platforms such as Fort Stewart, whose primary focus is to deploy units, experienced a more noticeable surge in TDA provider deployments when soldiers deployed, while Fort Bragg, which has missions other than deploying troops and a medical center instead of a community hospital, experienced little if any effect of deployments on TDA providers in-garrison at the local MTF.
There Was a Small Effect on Available Patient Care Hours When Soldiers Deployed
We observed a similarly weak relationship between soldier deployments and FTEs for civilian, contractor, and active-duty providers at MTFs. When 40 percent of the soldiers are deployed from an installation, we observed approximately 5 percent fewer outpatient-care FTEs from active-duty, civilian, and contractor providers available at the MTF than would be available if all soldiers were in-garrison. We did not observe a statistically significant change in the sum of patient-care FTEs and nonpatient-care FTEs at MTFs when soldiers deployed. In other words, when soldiers deployed, total provider FTEs at the MTFs did not change significantly, and outpatient-care FTEs decreased only slightly. We infer from these results that there is a weak relationship between patient-care FTEs recorded at Army MTFs and deployments.
Beneficiary Utilization of Health Care Across the Deployment Cycle
We examined the amount of health care utilized by beneficiaries (soldiers, families, retirees) across the deployment cycle. We looked at the mix of care from MTFs and civilian network providers utilized by beneficiaries to see whether there are changes during the deployment cycle.
We first describe the broader trends in Army beneficiary health care utilization before studying effects specifically related to the deployment cycle.
The Rate of Beneficiary Utilization of Health Care Increased 10 Percent over the Period Studied
Over the time period of the analysis, the rate of utilization grew approximately 10 percent in all categories: soldier and family member, MTF-enrolled or network-enrolled. On average, soldiers utilized health care resources at approximately twice the rate of family members. This finding is not surprising. Soldiers are required to visit a provider to receive permission to stay home from work and are not allowed to self-prescribe bed-rest for minor conditions, as civilians are. When soldiers attend sick call in an MTF clinic, these visits are recorded as utilization in the electronic medical record and appear in the data used in our analysis.
Families of soldiers at the installations in our analysis enrolled to both the MTFs and to civilian network providers for primary care. Families enrolled to the MTF got the majority of their care from the MTF, but still received 15 percent of their care from civilian providers. Family members enrolled to civilian providers received two-thirds of their care from civilian providers, and one-third from MTFs. We did not perform any analysis to understand differences in these populations, as it was outside the bounds of our analysis, but we did observe that family members enrolled to MTFs utilize 65 percent more care than family members enrolled to civilian providers.
Deployments Were Associated with an Increase in In-Garrison Soldier MTF Visits, but No Consistent Effect on Aggregate Utilization by Families
When soldiers deployed, soldier MTF visits decreased, although at a lesser rate than the decrease in soldier population in-garrison at the installation. When operational units deploy, we observed in our analysis that soldier visits decrease in one-third proportion to the deploying soldiers. For example, when 40 percent of the soldiers at an installation deploy, we observe that soldier visits to the MTF decrease by 13 percent. When soldiers deploy, soldiers who remain in-garrison use care at a higher rate than the overall soldier population.
We do not observe a statistically significant change in aggregate outpatient care at the MTFs by the populations of family members, or retirees and their dependents, when soldiers deploy.
Health Care Utilization Per-Provider Decreased When Soldiers Deploy
Although there was a slight decrease in the quantity of patient-care FTEs generated at MTFs when soldiers deploy, the rate of patient visits per provider FTE decreased during deployments. For example, when 40 percent of the soldiers from an installation are deployed, we observe outpatient workload per provider FTE to be approximately 5 percent lower across all types of visits.
Individual-Level Analysis of Family Health Care Utilization
We performed a longitudinal analysis of health care utilization by soldier family members. This analysis builds on a significant body of research defining how stress affects families, and how deployments affect family health care utilization. Our analysis extends the current body of research in several key ways: we analyzed care utilized by soldier family members at civilian providers as well as MTFs; we extended beyond regular outpatient care to look at ER and pharmacy utilization; we used statistical methods that control for other factors that may influence health care utilization; and we performed an analysis that looked at how the experience of individuals changed when a soldier deployed from their own family. Other studies that investigated effects of deployment on families have typically selected a short time horizon and compared the utilization by two distinct populations, those family members who have soldiers deploying, and those who do not. In our research we will be able to define the effects of deployment with more certainty, having controlled for other factors.
We studied the changes in health care utilization by soldier spouses. We saw that spouses decrease their utilization of outpatient care by 8 percent when soldiers in their family deploy. We were very interested to see that utilization at MTFs actually decreased by a greater amount, by 12 percent. Spouses changed their behaviors, and utilized more care from civilian providers, as outpatient care from civilian providers increased by 3 percent.
In our discussions with the Army medical community, they indicated their perception that family members tended to move from the area when soldiers in their families deployed. We designed an analysis to observe changes in outpatient care utilization from civilian providers who were located within 40 miles of the MTFs, and those located outside the area.***** We saw that while spouse utilization at civilian providers increased 3 percent overall, it actually decreased by 4 percent within 40 miles of the MTF, and increased by 35 percent at providers outside a 40-mile radius of the MTF. From this finding we conclude that spouses are indeed leaving the area when soldiers in their families deploy.
We studied mental health utilization by spouses. When soldiers in their families deployed, spouse utilization of total mental health care did not change much (increasing by 0.1 percent), but spouses were 4 percent more likely to visit for diagnoses related to mood, adjustment, and anxiety.****** These diagnoses are used commonly in health care research to define symptoms of stress and depression.
As we studied health care utilization by spouses when soldiers deployed, we performed a similar analysis of health care utilization by children. We saw that outpatient care for children did not change when soldiers deployed, but we saw a transfer in the source of care for children similar to what we saw for spouses. Children were 4 percent less likely to utilize outpatient care at MTFs, but were 1 percent more likely to utilize care at civilian providers.
We found the most dramatic changes in health care utilization when we studied children of custodial single parents.******* These children must stay with an alternate caregiver when their parent deploys. We saw that children of custodial single parents were 15 percent less likely to utilize outpatient care when their parents deployed. They were 26 percent less likely to utilize care at MTFs, and 13 percent more likely to utilize care at civilian providers.
When we studied pharmacy utilization by spouses and children, we saw that spouses were 7 percent less likely to use prescriptions when soldiers deployed, but they were 7 percent more likely to use prescriptions for antidepressants, which corresponds to our finding that spouses were more likely to utilize mental health care for stress- and depression-related diagnoses. Children were 1 percent more likely to utilize prescriptions, with increases in the likelihood of using antidepressants and anti-infectives.
Newer Army Families
In addition to studying all soldier family members enrolled in TRICARE Prime between 2004 and 2009, we performed an analysis of dependents of soldiers who joined the Army since 2001. It was our prediction that these families may experience greater deployment cycle effects, having less experience in dealing with family stresses associated with deployments and in accessing health care through TRICARE and Army MTFs.
Beneficiaries in post-2001 Army families exhibited deployment cycle effects consistent with the whole Army population, but with notably larger effects. When soldiers deployed, effects on these Army families were 50–75 percent larger for the following primary outcomes: a decrease in MTF visits by spouses, children, and children of single parents; and an increase in civilian visits outside the local catchment area.
Our study resulted in the following conclusions:
- Soldier utilization decreases in aggregate with deployments, but nondeploying soldiers use more care during these times.
- MTF capacity is not greatly affected when soldiers deploy. In aggregate, family member access does not appear impinged when soldiers deploy, and MTFs may be slightly less busy.
- The deployment cycle affects installations differently. The portion of soldiers that deploy from an installation and the portion of providers that deploy from the MTF are two factors that vary across installations and can affect changes in the demand for care and availability of appointments at the MTF.
- Spouses and children of single parents were less likely to utilize care when soldiers deployed and were noticeably more likely to utilize care outside their area.
- All categories of family members shifted their care from MTFs to civilian providers.
- Spouses and children utilized more mental health care for stress- and depression-related diagnoses when soldiers in the family deployed. Spouses also increased utilization of antidepressants.
- Decreases in MTF utilization and increases in civilian care outside the catchment area were even greater for younger Army families.
Eide, Matilda, Gregory Gorman, and Elizabeth Hisle-Gorman, “Effects of Parental Military Deployment on Pediatric Outpatient and Well-Child Visit Rates,” Pediatrics, Vol. 126, No. 1, July 1, 2010, pp. 22–27.
Gorman, Gregory H., Matilda Eide, and Elizabeth Hisle-Gorman, “Wartime Military Deployment and Increased Pediatric Mental and Behavioral Health Complaints,” Pediatrics, Vol. 126, No. 6, December 1, 2010, pp. 1058–1066.
McNulty, P.A., “Does Deployment Impact the Health Care Use of Military Families Stationed in Okinawa, Japan?” Military Medicine, Vol. 168, No. 6, June 2003, pp. 465–470.
* In the Army, deployable combat units are staffed by a greater proportion of junior enlisted soldiers than other types of units are, and so the installations in this analysis house a population of soldiers who are slightly younger and have slightly fewer dependents than the Army-wide demographic.
** As soldier populations grew at the installations over the course of the analysis, soldiers grew to represent a slightly larger percentage of the beneficiaries enrolled at the MTFs.
*** When soldiers deploy, they do not change their enrollment status. We identify when soldiers deploy through other elements in their personnel data designed for this purpose.
**** 40 percent of enrolled soldiers deploying is approximately the median effect at an installation in our analysis when a BCT or larger soldier population deploys.
***** We chose 40 miles as a definition of the area surrounding an MTF, since the military uses the same distance to define the population of beneficiaries to whom the MTF must be responsible for providing access to care.
****** During months when a soldier was not deployed, 48 percent of spouses utilized outpatient care, but only 6 percent of spouses utilized mental health care for the described diagnoses of mood, adjustment, and anxiety. So when we observed spouse utilization of mental health care for these diagnoses to increase relatively by 4 percent when soldiers deploy, in absolute terms the change was small.
******* In our analysis, 11 percent of the observations were of children of single, custodial parents. Children of two-parent households made up the predominance of observations in the analysis, and results for children of two-parent households are quite similar to those for the entire child population.
The research described in this article was sponsored by the United States Army and conducted by the RAND Arroyo Center.