UMERSE and Depression Screening
For this work researchers must identify and assess patients soon after a new depression treatment has been recommended. Each week they receive large data sets of patients age 60 and over who have been seen for depression in participating primary care clinics. This averages 425 individuals per week, but has ranged as high as 800. Patients' CareWeb (electronic medical record) documents must be screened for numerous inclusion and exclusion criteria prior to contacting their physician for permission to recruit patients, then followed by patient contact by mail and phone. Practically, patients must be screened within 1-2 days, a task that could require enormous manpower. With UMERSE, one individual is usually able to complete this task within the first day.
UMERSE has proven easy to learn, even for individuals with limited computer experience. The ability to create uniform search bundles eliminates the possibility of inconsistent screening of various patients. It also assists in protecting patient privacy in that no further examination of the charts of patients who are ineligible is necessary. Plans call for using UMERSE at the end of the study to conduct further chart abstractions for our enrolled patients.
UMERSE in the Clinical Trials Office
A leading academic research and Comprehensive Cancer Center's Clinical Trials Office (CTO) has over 80 staff members who have worked on hundreds of clinical trials. The group has found UMERSE to be invaluable for their work and have used it heavily over the last few years.
One of the clinical subjects coordinators uses UMERSE for data collection for a bone marrow transplant (BMT) national registry. She uses UMERSE to ensure that every clinic note is scanned for complications such as myocardial infarctions, cataracts, renal failure, and disease recurrences. UMERSE is also uses to search for infection prophylaxis medications used by the BMT patients.
Another clinical subjects coordinator in the area of hematology uses UMERSE to review hydrea doses as they fluctuate over time. These doses are dependent on the white blood count (WBC). As the WBC goes up, the dose goes up; As the WBC goes down, the dose goes down. One of the studies requested the highest dose the patients had received and the dates of hydrea administration. As these were outpatients, no medication administration pages were available. The doses of the drug were recorded only in the dictated clinic notes. For patients who had been in the study for over a year and a half, the prospect of finding the maximum dose was daunting. Fortunately, UMERSE accomplished this review in less than two minutes for three patients. UMERSE provided the hydrea listings and the ability to capture the highest dose in a matter of minutes. Without UMERSE, reviewing the records would have taken days.
A CTO clinical research coordinator has been using UMERSE for nearly 3 years. Using the tool she has been able to update important demographic and lifestyle information for a study that would have been difficult to obtain otherwise, such as smoking history. Prior to UMERSE she had to review every dictation, pathology report, and other records to find the few items needed, and often had to determine which speciality did the best dictations, such as radiation oncology or gynecology. But with UMERSE, she simply used terms such as "smoked" and other variations, and every notation in the patient's entire record that had any word would be flagged and displayed for review. Because UMERSE provides the context for which the terms are found it allowed her to ignore the ones that were obviously not needed. It has saved a tremendous amount of time.
UMERSE for Prior Authorizations
UMERSE and Lupus Screening
MiLES (the Michigan Lupus Epidemiology Surveillance Program) is a CDC-mandated, HIPAA-waived study being conducted at a leading academic research center in cooperation with the Michigan Department of Community Health. The purpose of the study is to determine the number of lupus patients who resided in Wayne and Washtenaw counties in the year 2002.
Lupus is a very complex rheumatologic disease which manifests itself in numerous ways and which may overlap with a number of other rheumatologic disorders. According to the American College of Rheumatology (ACR), there are 11 different diagnostic criteria which can manifest in lupus. A patient must have at least 3-4 criteria in order to warrant the diagnosis of lupus. In order to objectively determine this, it is necessary to review individual patient medical records and document the ACR criteria which patients have.
In order to ensure completeness in abstraction, all available electronic records are reviewed. The number of patient records can vary greatly- a patient may have one or two records or may have over 600.
Over 5000 patients at UM have been abstracted via Careweb, the University of Michign's electronic record system. Before UMERSE, it could take hours to abstract a patient. It was impossible to know if lupus was mentioned in a patient document unless the document was read. Often, the study team reviewed hundreds of records only to find no mention of lupus. With UMERSE, they investigators could search for any terms that were deemed important. Numerous search Bundles were set up, one just for the terms lupus and SLE, another for the ACR criteria for lupus. They were able to search for the descriptive terms used for lupus and for the medications used to treat lupus. They also developed a "miscellaneous" category in which they searched for individual terms that they were specifically looking for and may have missed with a manual review.
Using UMERE, they have now finished the abstraction at a leading academic research center. During the time period that they used UMERSE, they could often abstract 20 or more patients per day. Since that time, they have abstracted thousands of other patients at several other institutions using their electronic systems. Because UMERSE was not used at the other institutions, the data abstraction has taken far longer. Using UMERSE, they believe, could greatly benefit their reviewing records at those facilities in a timelier manner.
UMERSE and NCCN
The National Comprehensive Cancer Network's (NCCN) Oncology Outcomes Database Project is a multi-site data repository initiative that collects detailed information on the diagnosis, treatment and outcomes of patients with breast, ovarian, colon, and non-small cell lung cancer as well as patients with non-Hodgkin's lymphoma (NHL). The information gathered by the participating NCCN institutions is used to help shape the NCCN's cancer treatment guidelines that are seen as the benchmark for oncology treatment. The database also serves as a repository of data for investigators wishing to do research on various cancer treatment modalities and patient outcomes.
A leading academic research center currently participates in the breast, ovarian, non-small cell lung cancer, and NHL databases. Across all 4 disease sites, the University is currently collecting data on over 3000 patients. Patient data is abstracted at baseline, 6, and 12 month intervals and then annually until the patient's death. There are more than 100 data elements per patient that must be abstracted, many of which are very specific and can be difficult and time consuming to find.
Their data abstraction process, however, has been vastly improved with the use of UMERSE. Abstraction time for difficult to find data elements such as performance status, cancer staging information, medical events, and detailed drug therapy information has decreased drastically since using UMERSE. Team members were able to eliminate a greater than 3 year follow-up backlog of more than 1500 patients in only 6 months. The staff found UMERSE to be an essential tool for successful and efficient data abstraction. The ability to create tailored search bundles for specific chemotherapy agents and key words not only sped up the abstraction process, but also ensured high data quality.
One of the keys to UMERSE's usefulness is its user friendly design. One of the data managers had little familiarity with computers, but was able to learn how to use UMERSE with minimal guidance or assistance and has even trained new staff members on its use. One of the features of UMERSE that made it very user friendly is the alternative spelling and synonym expansion options that appear when entering enter a search term. This feature helps ensure that all the possible names for a specific drug or disease process are included in the search bundle allowing for increased accuracy in data abstraction. In addition, UMERSE allows for customization of document printing settings to highlight the specific search terms in the printed document, allowing staff to quickly find needed information even when the documents are printed out.
A leading academic research center's NCCN staff have found UMERSE to be an invaluable tool in the field of data abstraction. Not only did it drastically speed up the data abstraction process, but it is also helped to ensure high data quality and was extremely easy to use. They can't imagine going back to do doing data abstraction without it.
UMERSE and Plasma Biomarkers of Graft Versus Host Disease Validation
Bone Marrow Transplant investigators have used UMERSE as a tool in a NIH-funded program project grant "Project 3: Discovery and validation of plasma biomarkers of acute graft versus host disease" since July 2008.
Allogeneic hematopoietic cell transplantation (HCT) is a major therapy for cancers of the blood and bone marrow, including leukemias, lymphomas, and multiple myeloma. The use of HCT has increased as new techniques have been developed to allow for transplantation of patients who previously would not have been considered HCT candidates. Approximately 20,000 allogeneic HCTs were performed worldwide in 2008. However, the efficacy of allogeneic HCT has been impeded by frequent and severe graft versus host disease (GVHD).
The development of GVHD is tightly linked to the beneficial graft versus leukemia effect, the most potent form of immunotherapy. As of today, the diagnosis of acute GVHD is based solely on clinical criteria that may be confirmed by biopsy of one of the three target organs (skin, gastrointestinal (GI) tract, or liver). Currently no laboratory tests exist to diagnose GVHD or predict patient survival.
Using an antibody array, the BMT investigators recently reported a four biomarker panel in plasma of 428 individuals that discriminated between patients with and without GVHD and further predicted long term survival independent of GVHD severity (Paczesny et al., Blood 2009). However, this panel was not designed to be specific to any of the three target organs of GVHD. They are now planning, using a novel proteomic approach, to identify and validate biomarkers that can be used to build a biomarker panel for the diagnosis and prognosis of target specific GVHD. The ability to identify patients at high risk for GVHD early in their transplant and treatment course has important therapeutic consequences including more stringent monitoring and/or preventative care.
For this work the team must identify samples from cases and controls using a database that links samples to clinical information. Blood samples are collected weekly through the first post-HCT month, then monthly to day 100, at 6 months post-HCT, and then on the annual anniversary for 3 years post-HCT. In addition, samples are collected at the time of key clinical events, such as diagnosis of GVHD or development of symptoms consistent with GVHD (e.g., new rash, profuse diarrhea, elevated bilirubin), veno-occlusive disease (VOD), idiopathic pneumonia syndrome (IPS), sepsis, or relapse. Since 2000, a leading academic research center's BMT Program has created a research sample database repository containing more than 8500 samples obtained from over 850 individual allogeneic HCT recipients, including approximately 300 with GVHD.
As a standard procedure performed by the BMT program data managers, the clinical status at the time of sample collection is recorded and linked to the sample inventory for the major complications post-transplantation (GVHD, VOD, IPS, sepsis). However, the clinical data were not recorded for controls samples of patients developing, for instance, a non-GVHD rash. Thus, the team needed to screen these patients' Electronic Medical Record documents, a task that would have required precious research time if not for UMERSE.
With UMERSE, the team was able to identify 54 control samples obtained from patients with biopsy-proven non-GVHD rashes, which allowed for the testing of biomarker specificity for GVHD mediated skin damage. Similarly, the team was able to identify 74 patients with diarrhea symptoms and a GI tract biopsy negative for GVHD, which will allow testing of biomarker specificity for GVHD mediated GI damage. The screen was finalized in approximately one week, work that they estimate would have required two to three months without the tool. They found UMERSE to be very intuitive with the primary team member learning how to use it in 15 minutes. The ability to create uniform search bundles eliminated the possibility of inconsistencies in the screening of the patients. As a result, they are now in the process of submitting a second paper on plasma biomarkers of GVHD.
What they liked so much about UMERSE was that it saved precious time that could instead be used to advance the basic research. They say that they would have a hard time going back to the pre-UMERSE era.
UMERSE for Retrospective Reviews of Pediatric Surgical Patients
Investigators in a leading academic research center's Department of Surgery, Section of Pediatric Surgery have used UMERSE for several projects involving retrospective reviews of patients with congenital diaphragmatic hernia (CDH). This is a complex congenital defect that demands complex, lengthy care and often requires multiple operations. These projects gather a large range of data points for use in regression analyses. The goal is the identification of factors associated with survival in CDH, one of the most difficult problems in pediatric surgery today.
UMERSE greatly simplifies searching for data points like the size of chest tubes placed, complications experienced, dates of procedures, different materials used in surgical repair, etc. These data are reported in different places around the medical record, but UMERSE is able to search all the likely sources in one step. The interface is simple enough to learn quickly and allows for powerful searching very quickly.
UMERSE greatly simplifies the process of abstracting data from the medical record into a usable format. It also limits the information seen by the reviewer, which is not only a benefit for the reviewer's time but also helps protect patient privacy. It is an invaluable tool for detailed studies, and is useful for small and larger patient reviews alike. UMERSE greatly simplifies projects in which many data points must be collected on each patient and facilitates quality statistical analysis of these reviews.
UMERSE and Coding Compliance
The Coding Unit at a leading academic research center's Health System has been using UMERSE for medical record reviews since January 2009. Coding for inpatient hospital reimbursement is quite labor intensive as it requires review of hundreds of pages of documents every day. A leading academic research center's Coding Unit team must review 48,000 inpatient discharges a year. Many of those cases are long and involved and require our coding compliance staff to spend hours reading through physician progress notes on a single case. Ensuring that our coding and diagnosis related group (DRG) assignment is accurate and compliant is necessary for reimbursement, data collection and research purposes. UMERSE has been used by the Coding Compliance and Education Coordinators (CCEC) in identifying documentation to support code assignment. It is especially helpful on longer length of stay cases. The CCEC staff use UMERSE daily have often remarked on how easy it is to use and how much time it saves them in performing secondary quality reviews.
The Coding Unit's current plan is to train all of the coding and Clinical Documentation Specialist (CDS) staff on UMERSE. They believe that the bundling feature will prove beneficial to the CDS staff and they see a lot of potential for UMERSE to be a bridge to computer assisted coding.
UMERSE and Post-ERCP Pancreatitis
Researchers in the Department of Internal Medicine have been using UMERSE since 2007 for our project entitled "Alcohol Use and Cigarette Smoking as Risk Factors for Post-Endoscopic Retrograde Cholangiopancreatography Pancreatitis". Post-ERCP pancreatitis is a potential serious complication of all ERCP studies and occurs in up to 3.5% of people and 30-40% of high-risk patients. If clinicians can identify pre-procedurally those patients at higher risk for the condition, then they can offer risk-reducing strategies such as pre-procedural counseling and prophylactic pancreatic duct stenting.
The investigators used UMERSE to collect the necessary data and their study was recently published in Clinical Gastroenterology and Hepatology (Clin Gastroenterol Hepatol. 2009 Mar;7(3):353-8e4; PMID 19168153). Within the study publication, the investigators actually detailed how they used UMERSE (formerly EMERSE). Briefly, they utilized UMERSE to successfully, accurately, reproducibly, and efficiently search the institution's medical record to extract data on the various variables of interest. They were able to search over 7,000 records and extract all the data of interest within a matter of 6 weeks. Without UMERSE, this would have taken them over 18 months.
The researchers believe that UMERSE will absolutely need to be part of the next generation of medical record software and the electronic medical record.
UMERSE and Bone Marrow Transplantation
A leading academic research center has one of the leading bone marrow transplant (BMT) programs in the country. They are a leader not just in terms of the number of patients transplanted, but in terms of the basic science and clinical research generated. The BMT team now transplants over 240 patients per year. Approximately 40% of patients will develop lung complications after transplant, frequently in association with graft versus host disease. The use of bronchoscopy to identify various pathogenic organisms following transplant has been noted in a number of small case reports in the literature.
Using UMERSE, investigators recently analyzed the incidence of pulmonary complications in over 2000 allogenic and autologous stem cell transplants performed at our center from 2001-2009. With UMERSE, they were able to screen several hundred cases per day, with nearly all data fields completed. They were able to search for terms such as bronchoscopy, hypoxia, and pneumonia concurrently. Prior to UMERSE, they were only able to screen 20 to 30 charts per day, often without complete data being found. The researchers estimated that UMERSE saved them over 3 months of effort to gather the required data. As a result, the largest report of pulmonary complications and bronchoscopies post-transplant has now been generated. The results of this work were recently presented at the American Society of Bone Marrow Transplantation Conference. In addition, data from this effort is now being used as support for a federally funded cooperative group trial. Other centers have inquired how they were able to generate such a large data base so efficiently and the answer was clear -- through UMERSE.
The research team is now using UMERSE in multiple other clinical research projects throughout the transplant program. They are now using UMERSE to determine the incidence of thrombotic thrombocytopenic purpura (TTP), the incidence of BCNU related pneumonitis, and various risk factors for both chronic and acute graft versus host disease in the transplant patients. They anticipate that UMERSE will provide major contributions to the transplant program for years to come.
UMERSE for Neurosurgery Research
When a team of research scientists from a leading academic research center's Department of Neurosurgery wanted to initiate multiple studies, they turned to UMERSE. After using it for over one and a half years, they have found UMERSE to be an invaluable tool.
Using UMERSE has allowed them to search the entire patient record of a large volume of patients and to browse through large volumes of patient documentation quickly as a result of its key-term identification/highlighting function.
With the use of UMERSE, projects with a massive patient population have become more feasible, and projects evaluating rare diseases are now practical. Overall, UMERSE has significantly decreased work-time, increased productivity, and increased accuracy. It has brought a new dimension to clinical research. Its capacity to sift through massive volumes has proven invaluable to the study of large populations in the public health sector. The team is confident that UMERSE will revolutionize biomedical public health research.
UMERSE for the Childhood Cancer Survivorship Program
The Childhood Cancer Survivorship Program with the Department of Pediatric Hematology-Oncology is active in interventional research, aimed at reducing the long term burden of cancer treatment in childhood cancer survivors. We are engaged in research via a five institution consortium. Our goal is to develop interventional studies that can be completed in a relatively short time. If the study shows a significant outcome, we then propose the study to a much larger number of participants, within the pediatric cooperative groups.
We have used UMERSE to help construct study cohorts. For example, we are currently looking at potential cardiac intervention for patients who recieved high doses of anthracyclines as children. Our study cohort currently numbers 660; we were able to extract the precise number of patients who fit the eligibility of >250 mg/m2 anthracycline in less than 4 hours by using UMERSE. Had we to rely on the electronic medical record system note review, this collection would have taken significantly more time.
UMERSE is easy to use, and significantly decrease the time needed to extract clinical data from large numbers of patient records.
UMERSE and Infection Control
Infection Control and Epidemiology (ICE) at a leading academic research center's Health System constantly reviews the medical records of patients who have potentially been affected by healthcare associated infections (HAIs). This is to prevent unnecessary emotional and physical strain on the patients and families of the UMHS by minimizing the risk of future HAIs. Some familiar types of HAIs are surgical site infections, bloodstream infections, pneumonia, and urinary tract infections. Early HAI identification followed up with root-cause analysis is crucial to infection prevention and process improvement. Reviewing medical records for signs and symptoms of infection can be labor intensive, requiring hundreds of man hours per month. ICE is currently using UMERSE to quickly identify infections in the surgical population.
Since utilization of the search engine began, ICE has more than doubled the amount of patient records reviewed for surgical site infections. This includes surveillance based on expanding Centers for Medicare & Medicaid Services (CMS) reimbursement guidelines as well as per physician request. In a recent comparison, UMERSE had 100% specificity and a 94.8% sensitivity rate for surgical site infection identification as compared with the previously used manual search method. Because UMERSE can be so specific it has decreased the time necessary to perform surgical site infection surveillance while increasing the number of medical records reviewed; an impressive time savings of an estimated 40-60 working hours per month.
Additional process improvement benefits from UMERSE include a decreased dependence on outside departments for information. For example, it is no longer necessary to wait for a report from the hospital's microbiology lab; UMERSE can run the search for you. ICE also uses UMERSE to assist in the identification of patients with central venous catheters, the incidence of multi-drug resistant organism colonization/infection throughout our patient populations, exploration of potential infection clusters and for outbreak investigations. Because of UMERSE, ICE has become a more efficient and effective department within UMHS.
UMERSE and Child Behavioral Health
Investigators in Child Behavioral Health at a leading academic research center have been working to identify an unbiased sample of patients with disorders of sex development (DSD) at 4 investigation sites: a leading academic research center, Children's Hospital of Philadelphia, Albany Medical College, and the University of Miami. To circumvent the potential for selection bias introduced by case ascertainment relying on idiosyncratic search algorithms, site co-investigators were supplied with a list of ICD-9 diagnostic codes to guide their search. Electronic systems used for billing or record keeping purposes at each site were used to construct a preliminary list of patients. This list was then cross-matched with demographic variables (e.g., patient date of birth) which are also often stored in such electronic databases.
Following this strategy, they quickly understood that ICD-9 diagnostic codes combined with demographic information could not, without extensive review of dictated notes, deliver a sample frame of eligible patients. For example, the ICD-9 code 752.61 (hypospadias) returned a vast number of 46,XY patients, only a subset of whom would be categorized as DSD according to the consensus statement that introduced the new nomenclature. In another example, they were uncertain whether a series of other ICD-9 codes would be used at all by providers for patients classified as DSD, including 752.4 (Abnormalities of cervix, vagina, and external female genitalia), 752.40 (Unspecified anomaly of cervix, vagina, and external female genitalia), 752.49 (Other anomalies of cervix, vagina, and external female genitalia), or 752.69 (Other penile anomalies). The process of culling the substantial number of, ultimately, ineligible cases was immensely time-consuming.
The investigators thus learned that although the ICD-9 codes helped narrow the pool of possible candidate patients, these codes were crude in that they did not accurately capture the nuances of a patient's diagnosis. Appreciating how labor intensive this stage of the research plan for the study ultimately turned out to be the principle investigator at a leading academic research center turned to UMERSE to search through the dictated notes and other electronic medical record (EMR) elements of those patients identified using a combination of ICD-9 codes and demographic qualifiers. The use of EMERSE validated the accuracy of research staff identified cases by generating an almost identical list of patients. It accomplished this task, however, in approximately one third of the time.