Collaborative Project to Improve Detection Rates of Hereditary Hemorrhagic Telangiectasia in the KPNC Population
The Kaiser Foundation Research Institute, a division of Kaiser Foundation Hospitals, a California nonprofit public benefit corporation, has been awarded a $35,000 grant for research that will addressed the question: “What is the accuracy of specific combinations of ICD-9 codes in the diagnosis of Hereditary Hemorrhagic Telangiectasia (HHT)?”
HHT is under-diagnosed because physicians can fail to connect HHT’s diverse manifestations (e.g. epistaxis, skin telangiectasias, anemia) as part of an underlying syndrome. Researchers at the Centers for Disease Control (CDC) have developed combinations of ICD-9 codes, the “HHT Algorithms” (HHTA), which appear to have good accuracy in identifying HHT cases. The research team at the Kaiser Foundation Research Institute, led by Jonathan Zaroff, MD, with collaborating investigators from the CDC, the University of California at San Francisco, and Dr. Marie Faughnan at the University of Toronto, believes that the 3 million members of Kaiser Permanente Northern California (“KPNC”) represents an appropriate population to use to validate and further refine the HHTA and allow for previously unrecognized HHT cases to benefit from guidelines-based HHT care.
- Aim 1: Quantify the number of KPNC members without a prior HHT diagnosis who meet the specific coding criteria for the HHTA; this will be done by the KP Division of Research database programmer.
- Aim 2: Assess the validity of the HHTA using chart review of EMR data (clinical notes, lab, imaging) to determine the presence or absence of selected clinical diagnostic criteria for HHT in all potential cases identified by the programmer.
- Aim 3: Validate the HHTA by contacting a subgroup of the identified KPNC members and performing a comprehensive clinical assessment for HHT.
This study should enhance current knowledge regarding the usage of ICD-9 code based searches in identifying previously undiagnosed people with HHT as well as the yield of a comprehensive clinical assessment for HHT in this population. This is a twelve month study that began in March 2013.
Research Study Update
Aim 1 – A total of 3,065,210 KPNC members’ records were queried: (age <35) 528,030; (age 36-50) 652,319; (age >50) 1,124,441. The total number of KPNC members with known HHT (having an existing ICD-9 code of 448.0) and the number of members with codes meeting criteria for each of the four HHT algorithms was calculated along with the prevalence among the total population and in specific age groups. The Aim 1 results indicate that the prevalence of diagnosed HHT within the KPNC population, 0.42 per 10,000 or approximately 1 per 24,000, is substantially lower than the likely true prevalence of HHT (1 in 5,000-8,000). Thus it is very likely that there are KPNC members with undiagnosed HHT.
Aim 2 – The data review focused on the elements of the algorithms as well as the clinical diagnostic (Curacao) criteria for HHT. It was unknown prior to this project phase to what extent some of the Curacao criteria (epistaxis, telangiectasia, family history of HHT) would be recorded in the EMR. This aim was achieved by performing chart review of a sample of members identified as possibly having HHT by the HHTA. The Aim 2 study population was substantially older than the general KPNC population, with a mean age of 68 (median 71) and a high mortality rate (9%) since the date of programming (3/1/2013). In one case, chart review revealed a known diagnosis of HHT that had not been coded in the EMR and it is likely that the low prevalence of diagnosed HHT observed in the KPNC population is at least partially due to under-coding.
Aim 3 – Thirty-seven patients were identified during chart review as having a reasonable likelihood of having HHT. These patients were invited to visit Kaiser San Francisco Medical Center for the Aim 3 study visits, which include an interview and physical exam by the PI, and interview with a Genetics Counselor (co-investigator), and lab testing that includes contrast echocardiography, brain MRI, iron studies, and a complete blood count (when not already done as part of patients’ clinical care at KPNC).