RPMS Surveillance Capacity Project
Background
The process of monitoring the rates and trends of various diseases is referred to as surveillance. Gaining the ability to identify these rates and trends is vital to improve health outcomes within communities.
Within American Indian and Alaska Native (AI/AN) communities, there are many sources of information that can potentially be used to improve surveillance strategies. One of these sources of information is called Resource and Patient Management System (RPMS).
What is Resource and Patient Management System (RPMS)?
RPMS is a computerized Health Information System that has been used by most Indian health care programs since the 1980s. RPMS is a comprehensive suite of packages with many outstanding features that are useful for the daily management of patients.
Although RPMS has a vast amount of data on the health status of individual patients, a detailed analysis of its usefulness and accuracy for surveillance has not been done. The EpiCenter has recently been invited by the national IHS Epidemiology Program to enter into a contract to do such an analysis.
Project Goal
o To perform an analysis of the RPMS system from a national sample of Title-I 638, Title-III Self Governance and IHS health care facility programs, to determine the accuracy and reliabity of this system as a potential resource for epidemiologic surveillance.
Objectives
o To determine the extent that RPMS can be utilized as an accurate and reliable source of information for epidemiolgic surveillance on a variety of health conditions within Indian communities throughout the nation.
o To provide a set of recommendations that will improve the capacity of RPMS.
Methods
This project will compare retrospective data of medical diagnosis recorded in RPMS and other sources of medical information such as lab records documented by selected Indian health care programs. This information will be used to determine if these diagnosed cases correspond to the case definitions provided by the Centers for Disease Control and Prevention (CDC).
Duration of the Project
This project has been funded for one year and is expected to be completed by late December 1998.
Review the results of this project:
National Evaluation of RPMS
A National Evaluation of the Indian Health Service’s
Resource Patient Management System (RPMS)
Northwest Portland Area Indian Health Board
Northwest Tribal Epidemiology Center
Kelly Gonzales, MPH
Project Background
Resource Patient Management System (RPMS) is a computerized health information system used by almost all Indian health care programs since 1984. RPMS is a comprehensive suite of packages with many features including a module called Patient Care Component (PCC) which is particularly useful in the daily management of patients. The PCC package is used for storing information from the medical records for case management and reimbursement.
Limited Information
Since RPMS holds a vast amount of data on the health status of individual patients, this system can potentially serve as a resource providing data for surveillance on a number of health conditions. However, other than limited studies of specific diseases, there have been no systematic studies conducted to determine the capacity of RPMS for disease surveillance.
Project Scope
In 1997 the Northwest Portland Area Indian Health Board (NPAIHB) contracted with the Indian Health Service Headquarters West, Epidemiology Program to perform a nation-wide assessment of RPMS to determine the accuracy of data stored in the decentralized RPMS health information system. A systematic review of medical records and laboratory data enabled project staff to evaluate the extent that diagnoses entered in RPMS corresponded with published case definitions.
Methodology
Selection Strategies for Indian Health Care Programs
Using a stratified sampling design, a cross-sectional sample of forty-four Indian health care programs from each of the IHS Areas.
Number of medical charts reviewed by IHS Area:
Oklahoma 602
Aberdeen 329
Alaska 908
California 423
Nashville 94
Phoenix 113
Portland 512
Total 2,982
Due to funding limitations, reviews were completed at twenty-two programs located in seven IHS Areas including:
Alaska, Aberdeen, California, Nashville, Oklahoma, Phoenix and Portland Area.
For selection purposes, all Indian health care programs were stratified into three different categories distinguished by operating authority:
1) IHS
2) 638 Contract
3) Self Governance Compact and a sample of Indian health care programs were randomly select using a locally developed random computer program.
Selection Strategies for Diagnosis
Participating programs provided the Project Director with access to their local RPMS system. The following strategy was used to search for a list of patients entered in RPMS with a particular diagnosis:
Visits with a particular study diagnosis using the ICD-9 code with a visit between 0100195 through 123197 were searched using the RPMS Q-man query utility (fileman print option). This list was displayed to exhibit the ICD-9 code; Provider narrative; date of visit; and date the visit was created.
Using stratified random sampling methods, a sample of medical records was selected at each participating programs.
To obtain a statistically reliable estimate of medical records, project staff referred to the Sample Size Calculations table found in the 1998 Indian Health diabetes Mellitus Services Chart Audit for Quality Assurance and Quality Improvement.
The number of medical records reviewed was based on the total number of diagnoses that was generated in RPMS for a particular study diagnosis. For example, a search was conducted for Hepatitis A and a list of patients with the particular search criteria was generated.
A sample of medical records corresponding to this list was then randomly selected. This selection process occurred for each diagnosis included in this project.
|
Analysis
A total of 2,982 medical records were reviewed for the twenty-two Indian health care programs combined.
Of the diagnoses that did not meet CDC case definitions, 20% were documented in RPMS for Chlamydia; 17% for STD; 14% for Diabetes Mellitus; 11% for Gonorrhea; 9% for PID; 8% for Shigellosis; 8% for Other; 7% for TB (active; and 5% for Pertussis.
Of the diagnoses that were least likely to meet case definitions, 42% were determined to be Rule Out diagnoses where the entries documented by the medical provider subsequently had no additional confirmatory information or there wasw a negative laboratory result for the particular study diagnosis; 40% contained insufficient documentation; 15% did not undergo a laboratory test to confirm the disease, however had contact with a person who had been diagnosed with the disease; and 2% were either an immunization, STD evaluation or education.
Conclusion and Recommendations
Upon completing a systematic review of medical records at participating Indian health care programs, there was substantiating evidence in 77% of the medical records to confirm the corresponding diagnosis according to published case definitions.
Based on the results of this project, it is recommended that considerations should be made to facilitate certain improvements in the quality of RPMS data.
The following is a list of recommendations for the improvement of RPMS data processes. These recommendations are based on evidence collected during site visits for this project.
1. Begin communication process between staff involved with data input, including medical providers, data entry personnel, nursing staff, CHRs, and administration, to discuss methods that are appropriate for the program to ensure data improvement.
2. Collectively work to “clean up” medical records. This process can be shared by many staff.
3. Create preformed PCC forms that serve as a reminder to medical staff which procedures need to be conducted (i.e., evaluation of tobacco use) and indicates the correct name for the diagnosis (i.e., type 1 diabetes vs. type 2 diabetes instead of indicating only dm or diabetes on the PCC form).
4. Rule outs should not be encoded in RPMS as a diagnosis. Instead only symptoms should be encoded.
Additional Information
Many of the sites visited for this project experienced other factors that influenced the quality of data found in RPMS, such as understaffing, limited resources for formal training of data entry staff and frequent changes in medical staff and support staff. Designing strategies to address these factors are much more challenging and should be developed on systematic factors that are unique to the individual Indian health care program.
Kelly Gonzales, MPH
Surveillance Capacity Project Director
