Data and Methods for Patient and Provider-Level Analyses

Appendix A

Data and Methods for Patient and Provider-Level Analyses: UNPROFOR and Somalia

UNPROFOR Data and Methods

For the patient utilization analyses presented in Chapter Three, we drew from several different data sources. For the two Army Mobile Army Surgical Hospitals (212th MASH and 502nd MASH), we were able to obtain patient-level data from the Directorate of Patient Administration Systems and Biostatistics Activities (PASBA), AMEDD Center and School. These data covered the period between November 1992 and October 1993. Data elements included number of outpatient visits, number of admissions, and length of stay for each patient category.

Since our interest was also in comparing differences in utilization between U.S. personnel, foreign military, foreign civilians, and UN/NATO employees and officers, for this analysis we further grouped the Army patient data into these four categories. Assumptions made in determining these groupings were similar to those described below for the Somalia deployment. The only difference in patient groupings between Somalia versus UNPROFOR and Provide Promise is that for the latter deployment, UN personnel and NATO employees and officers were combined into a single category.

To examine overall differences in utilization across the four rotations, we present patient-level data obtained from briefing charts put together by the Navy's Fleet Hospital 6, which summarized for each rotation the number of outpatient visits, number of admissions, and proportion of inpatients with disease versus trauma-related conditions. For this comparison we needed information on both the Air Force and Navy hospitals that undertook the third and fourth rotations into Zagreb, in addition to those rotations done by the U.S. Army. This information could only be obtained from briefing charts. In addition, we utilized unpublished length of stay data for the Navy's Fleet Hospital 6 obtained from CAPT James Carlisle, Chief of Clinical Services, in order to compare how average length of stay differed across contingents during the fourth rotation. We used data for the Navy hospital for this comparison because the PASBA data for the two Army hospitals did not allow us to break down the foreign military category by individual contingent.

Somalia Data and Methods

Provider-Level Analysis

To track medical support for the mission in Somalia, we use data on Joint Task Force-Somalia (JTF-S) Professional Fillers (PROFIS) personnel. These data do not include the organic medical assets belonging to the 10th Mountain Division (the division that served as the backbone of U.S. forces in Somalia). The organic medical assets of a field unit or division typically include physician assistants (PAs), medics, and Medical Service Corps (MSC) officers who are regularly assigned to a medical field unit on a full-time basis.[1] Physicians, nurses, and other specialties will mostly be designated as PROFIS personnel, with these individuals spending most of their time, when not deployed, in a fixed facility. Therefore, although the PROFIS personnel included in this analysis represent the bulk of the medical support in Somalia, they do not represent all of it, since a few individuals, physicians for example, were assigned full time to the 10th Mountain Division as part of the division's organic medical assets.

Since AMEDD personnel came and went at various times during the Somalia deployment, we counted the number of PROFIS personnel at the midpoint of each month. This allowed us to obtain a consistent snapshot of what the medical support looked like for each month of the deployment and how it changed over time.

We began with data on the total number of PROFIS personnel deployed to Somalia by unit and area of concentration (AOC) occupation. The number of PROFIS personnel at the midpoint of each month is shown over the course of the entire operation, starting with November 1992 through March 1994. Certain AOCs were grouped as follows: under the Preventive Medicine Officer category, preventive medicine physicians, entomologists, and environmental science officers were grouped. Administrative, logisticians, and operations officers were included under the health services officer category (i.e., those officers involved in administration or the operational aspects of a medical field unit).[2] The behavioral sciences category included psychologists and social workers.

For this analysis, because we wanted to examine how the specialty mix changed over time, we further grouped the AOCs into the following specialty categories: preventive medicine, primary care and medicine, surgical and related specialties, mental health, dental, nursing, administration and health services, and other specialties. Table A.1 lists the specialty categories and the AOCs that fall within each grouping.

Table A.1
Specialty Categories and AOCs--Somalia

Specialty AOC
Preventive Medicine
Preventive Medicine Officer 60C
Community Health Nurse 66B
Preventive Medicine Officer 67C
Primary Care and Medicine
Pediatrician/Internist 60P,61F
Family Practitioner/ER/PA 61H,62A,65D
Flight/Field Surgeon 61N,62B
Surgical and Related
Ob/Gyn 60J
General/Thoracic/Orthopedic 61J,61K,61M
Other Surgical Specialties 60N,60S,60T,60Z,66F
Mental Health
Psychiatrist 60W
Mental Health Nurse 66C
Behavioral Sciences Officer 67D
General Dentist 63A
Comprehensive Dentist 63B
Oral Surgeon 63N
Operating Room RN 66E
Medical-Surgical/Clinical RN 66H,66J
Administrative and Health Services
Executive Medicine Officer 60A
Medical Maintenance Officer 670A
Health Services Officer 67A
Aeromedical Evac Officer 67J
Pulmonologist 60F
Infectious Disease Officer 61G
Diagnostic Radiologist 61R
Dietitian 65C
Laboratory Sciences Officer 67B
Pharmacy Officer 67E
Optometrist 67F

Interpretation of the AOCs or specialty mix has to be done with caution. While a PROFIS individual is designated to fill a specific slot in a deploying medical unit, his or her MTF commander has a fair amount of leeway in terms of who actually may be deployed. If a commander cannot afford to lose a particular individual, he may send instead another to fill the PROFIS slot. In addition, the AMEDD has recently revised some of its AOC codes, which has made the interpretation of some of the PROFIS taskings ambiguous.

Patient-Level Analysis

To examine changes in patient utilization over the course of the Somalia mission, we obtained patient-level data from the Patient Administration Systems and Biostatistics Activities, MEDCOM. The in-theater data had been collected by the individual hospital units and then reported to the MEDCOM.

The patient data cover the period between January 1993 and January 1994. Note that the data do not cover the initial few months (November and December 1993) or the latter few months (February and March 1994) of this operation. This is in contrast to the provider data, which covered all 17 months of this deployment. Also note that patient data were unavailable for April 1993 and August 1993, the two months when the rotation of U.S. troops and hospital units into the theater took place.

We were able to obtain data on outpatient visits, admissions, length of stay, clinic of service (or disposition), and patient category. Since our interest was in comparing differences in utilization between U.S. personnel, foreign military, and foreign civilians, for this analysis we further grouped patients into these three categories, as shown in Table A.2.

In terms of the foreign civilians and other foreign nationals, the three hospital units were not consistent in how they coded patients between these two categories, so we combined them into a single category. The foreign civilian category we use consisted mostly of Somali nationals, with a few non-U.S. civilians (e.g., relief workers).

We were also interested in comparing the distribution of patients across clinical services within a hospital. Table A.3 shows the four groupings we used to examine the inpatient services and lists the type of visits categorized under the outpatient listing. We were unable to separate out emergency room visits from other types of outpatient visits, since the coding was inconsistent across the three types of hospital units. For a separate analysis we also compared the distribution of surgical patients across the different surgical specialties, using data on patients' disposition (i.e., the last clinic of service) (see Table A.3).

Table A.2
Patient and Clinical Services Categories--Somalia

U.S. Personnel
  Army active duty
  Navy active duty
  Marine active duty
  Air Force active duty

  Federal department employee
  Federal agency employee
  Dependent, non-DoD federal agency
  Contract employee

Foreign military

NATO military personnel
Non-NATO military personnel
Foreign civilians
Foreign civilian
Other foreign nationals

Table A.3

Clinical Services and Surgical Disposition Categories

Clinical Services Surgical Disposition
Inpatient admissions
  Internal medicine
   Internal medicine
   Infectious diseases
   Family practice medicine


   Family practice obstetrics

   Family practice psychiatry

Outpatient visits
  Emergency room
  Orthopedic outpatient/casts
  Mental health/social worker
  Primary care/family practice
  Acute minor illness visit


Other surgery
  Head and neck

[1] Physician assistants (PAs) are assigned to a medical unit attached to a division full time, in theory. In actuality, because they also need to see patients, they will spend some of their time in a fixed facility.

[2] There are few PROFIS individuals in health services officer slots, since usually these AOCs will be organic to the division and the medical unit. Therefore, it is less common to augment these administrative-type positions via PROFIS.

Previous chapter
Appendix B