Outcomes of patient self-referral for the diagnosis of several rare inherited kidney diseases

PurposeTo evaluate self-referral from the Internet for genetic diagnosis of several rare inherited kidney diseases.MethodsRetrospective study from 1996 to 2017 analyzing data from an academic referral center specializing in autosomal dominant tubulointerstitial kidney disease (ADTKD). Individuals were referred by academic health-care providers (HCPs) nonacademic HCPs, or directly by patients/families.ResultsOver 21 years, there were 665 referrals, with 176 (27%) directly from families, 269 (40%) from academic HCPs, and 220 (33%) from nonacademic HCPs. Forty-two (24%) direct family referrals had positive genetic testing versus 73 (27%) families from academic HCPs and 55 (25%) from nonacademic HCPs (P = 0.72). Ninety-nine percent of direct family contacts were white and resided in zip code locations with a mean median income of $77,316 ± 34,014 versus US median income $49,445.ConclusionUndiagnosed families with Internet access bypassed their physicians and established direct contact with an academic center specializing in inherited kidney disease to achieve a diagnosis. Twenty-five percent of all families diagnosed with ADTKD were the result of direct family referral and would otherwise have been undiagnosed. If patients suspect a rare disorder that is undiagnosed by their physicians, actively pursuing self-diagnosis using the Internet can be successful. Centers interested in rare disorders should consider improving direct access to families.     

The utility of the traditional medical genetics diagnostic evaluation in the context of next-generation sequencing for undiagnosed genetic disorders

PurposeThe purpose of this study was to assess the diagnostic yield of the traditional, comprehensive clinical evaluation and targeted genetic testing, within a general genetics clinic. These data are critically needed to develop clinically and economically grounded diagnostic algorithms that consider presenting phenotype, traditional genetics testing, and the emerging role of next-generation sequencing (whole-exome/genome sequencing).MethodsWe retrospectively analyzed a cohort of 500 unselected consecutive patients who received traditional genetic diagnostic evaluations at a tertiary medical center. We calculated the diagnosis rate, number of visits to diagnosis, genetic tests, and the cost of testing.ResultsThirty-nine patients were determined to not have a genetic disorder; 212 of the remaining 461 (46%) received a genetic diagnosis, and 72% of these were diagnosed on the first visit. The cost per subsequent successful genetic diagnosis was estimated at $25,000.ConclusionAlmost half of the patients were diagnosed using the traditional approach, most at the initial visit. For those remaining undiagnosed, next-generation sequencing may be clinically and economically beneficial. Estimating a 50% success rate for next-generation sequencing in undiagnosed genetic disorders, its application after the first clinical visit could result in a higher rate of genetic diagnosis at a considerable cost savings per successful diagnosis.Genet Med16 2, 176–182.     

Unsupervised ensemble ranking of terms in electronic health record notes based on their importance to patients

BackgroundAllowing patients to access their own electronic health record (EHR) notes through online patient portals has the potential to improve patient-centered care. However, EHR notes contain abundant medical jargon that can be difficult for patients to comprehend. One way to help patients is to reduce information overload and help them focus on medical terms that matter most to them. Targeted education can then be developed to improve patient EHR comprehension and the quality of care.ObjectiveThe aim of this work was to develop FIT (Finding Important Terms for patients), an unsupervised natural language processing (NLP) system that ranks medical terms in EHR notes based on their importance to patients.MethodsWe built FIT on a new unsupervised ensemble ranking model derived from the biased random walk algorithm to combine heterogeneous information resources for ranking candidate terms from each EHR note. Specifically, FIT integrates four single views (rankers) for term importance: patient use of medical concepts, document-level term salience, word co-occurrence based term relatedness, and topic coherence. It also incorporates partial information of term importance as conveyed by terms’ unfamiliarity levels and semantic types. We evaluated FIT on 90 expert-annotated EHR notes and used the four single-view rankers as baselines. In addition, we implemented three benchmark unsupervised ensemble ranking methods as strong baselines.ResultsFIT achieved 0.885 AUC-ROC for ranking candidate terms from EHR notes to identify important terms. When including term identification, the performance of FIT for identifying important terms from EHR notes was 0.813 AUC-ROC. Both performance scores significantly exceeded the corresponding scores from the four single rankers (P < 0.001). FIT also outperformed the three ensemble rankers for most metrics. Its performance is relatively insensitive to its parameter.ConclusionsFIT can automatically identify EHR terms important to patients. It may help develop future interventions to improve quality of care. By using unsupervised learning as well as a robust and flexible framework for information fusion, FIT can be readily applied to other domains and applications.     

Electronic health record acceptance by physicians: Testing an integrated theoretical model

ObjectiveSeveral countries are in the process of implementing an Electronic Health Record (EHR), but limited physicians’ acceptance of this technology presents a serious threat to its successful implementation. The aim of this study was to identify the main determinants of physician acceptance of EHR in a sample of general practitioners and specialists of the Province of Quebec (Canada).MethodsWe sent an electronic questionnaire to physician members of the Quebec Medical Association. We tested four theoretical models (Technology acceptance model (TAM), Extended TAM, Psychosocial Model, and Integrated Model) using path analysis and multiple linear regression analysis in order to identify the main determinants of physicians’ intention to use the EHR. We evaluated the modifying effect of sociodemographic characteristics using multi-group analysis of structural weights invariance.ResultsA total of 157 questionnaires were returned. The four models performed well and explained between 44% and 55% of the variance in physicians’ intention to use the EHR. The Integrated model performed the best and showed that perceived ease of use, professional norm, social norm, and demonstrability of the results are the strongest predictors of physicians’ intention to use the EHR. Age, gender, previous experience and specialty modified the association between those determinants and intention.ConclusionsThe proposed integrated theoretical model is useful in identifying which factors could motivate physicians from different backgrounds to use the EHR. Physicians who perceive the EHR to be easy to use, coherent with their professional norms, supported by their peers and patients, and able to demonstrate tangible results are more likely to accept this technology. Age, gender, specialty and experience should also be taken into account when developing EHR implementation strategies targeting physicians.     

Impact of integrating genomic data into the electronic health record on genetics care delivery

PurposeIntegrating genomic data into the electronic health record (EHR) is key for optimally delivering genomic medicine.MethodsThe PennChart Genomics Initiative (PGI) at the University of Pennsylvania is a multidisciplinary collaborative that has successfully linked orders and results from genetic testing laboratories with discrete genetic data in the EHR. We quantified the use of the genomic data within the EHR, performed a time study with genetic counselors, and conducted key informant interviews with PGI members to evaluate the effect of the PGI’s efforts on genetics care delivery.ResultsThe PGI has interfaced with 4 genetic testing laboratories, resulting in the creation of 420 unique computerized genetic testing orders that have been used 4073 times to date. In a time study of 96 genetic testing activities, EHR use was associated with significant reductions in time spent ordering (2 vs 8 minutes, P < .001) and managing (1 vs 5 minutes, P < .001) genetic results compared with the use of online laboratory-specific portals. In key informant interviews, multidisciplinary collaboration and institutional buy-in were identified as key ingredients for the PGI’s success.ConclusionThe PGI’s efforts to integrate genomic medicine into the EHR have substantially streamlined the delivery of genomic medicine.     

Phenotypic presentation of Mendelian disease across the diagnostic trajectory in electronic health records

PurposeTo investigate the phenotypic presentation of Mendelian disease across the diagnostic trajectory in the electronic health record (EHR).MethodsWe applied a conceptual model to delineate the diagnostic trajectory of Mendelian disease to the EHRs of patients affected by 1 of 9 Mendelian diseases. We assessed data availability and phenotype ascertainment across the diagnostic trajectory using phenotype risk scores and validated our findings via chart review of patients with hereditary connective tissue disorders.ResultsWe identified 896 individuals with genetically confirmed diagnoses, 216 (24%) of whom had fully ascertained diagnostic trajectories. Phenotype risk scores increased following clinical suspicion and diagnosis (P < 1 × 10⁻⁴, Wilcoxon rank sum test). We found that of all International Classification of Disease–based phenotypes in the EHR, 66% were recorded after clinical suspicion, and manual chart review yielded consistent results.ConclusionUsing a novel conceptual model to study the diagnostic trajectory of genetic disease in the EHR, we demonstrated that phenotype ascertainment is, in large part, driven by the clinical examinations and studies prompted by clinical suspicion of a genetic disease, a process we term diagnostic convergence. Algorithms designed to detect undiagnosed genetic disease should consider censoring EHR data at the first date of clinical suspicion to avoid data leakage.     

Toward a model of shared meaningful diagnosis

ObjectivesThe purpose was to explain the process of diabetes-related diagnosis that prompts patient action (behavior change or treatment adherence). A secondary purpose was to identify barriers/gaps that prevent those outcomes.MethodsUsing a grounded theory approach, we explored diagnosis from the patient’s perspective and through the lens of the electronic health record (EHR). A thematic analysis was conducted on interview and EHR data from 28 patients, using the constant comparative method.ResultsThe emerging model of shared meaningful diagnosis included four stages: stimulus to screen, medical decision making, medical information transfer, and patient sensemaking. Barriers to a meaningful diagnosis emerged in clinical documentation, clinician communication, and patient sensemaking.ConclusionsThis study expands current understanding of “diagnosis,” suggesting additional stages between diagnostic labeling and disease management. The additional stages of medical information transfer and patient sensemaking are critical steps to a shared meaningful diagnosis that could enable teamwork among the patient and healthcare team.Practice ImplicationsTo sustain meaningful diagnosis for the patient, clinicians should document what language they used to explain the diagnosis to the patient so that subsequent clinicians can use similar language. Clinicians who work as a team should unify their approach to discussing prediabetes.     

Relational machine learning for electronic health record-driven phenotyping

ObjectiveElectronic health records (EHR) offer medical and pharmacogenomics research unprecedented opportunities to identify and classify patients at risk. EHRs are collections of highly inter-dependent records that include biological, anatomical, physiological, and behavioral observations. They comprise a patient’s clinical phenome, where each patient has thousands of date-stamped records distributed across many relational tables. Development of EHR computer-based phenotyping algorithms require time and medical insight from clinical experts, who most often can only review a small patient subset representative of the total EHR records, to identify phenotype features. In this research we evaluate whether relational machine learning (ML) using inductive logic programming (ILP) can contribute to addressing these issues as a viable approach for EHR-based phenotyping.MethodsTwo relational learning ILP approaches and three well-known WEKA (Waikato Environment for Knowledge Analysis) implementations of non-relational approaches (PART, J48, and JRIP) were used to develop models for nine phenotypes. International Classification of Diseases, Ninth Revision (ICD-9) coded EHR data were used to select training cohorts for the development of each phenotypic model. Accuracy, precision, recall, F-Measure, and Area Under the Receiver Operating Characteristic (AUROC) curve statistics were measured for each phenotypic model based on independent manually verified test cohorts. A two-sided binomial distribution test (sign test) compared the five ML approaches across phenotypes for statistical significance.ResultsWe developed an approach to automatically label training examples using ICD-9 diagnosis codes for the ML approaches being evaluated. Nine phenotypic models for each ML approach were evaluated, resulting in better overall model performance in AUROC using ILP when compared to PART (p = 0.039), J48 (p = 0.003) and JRIP (p = 0.003).DiscussionILP has the potential to improve phenotyping by independently delivering clinically expert interpretable rules for phenotype definitions, or intuitive phenotypes to assist experts.ConclusionRelational learning using ILP offers a viable approach to EHR-driven phenotyping.     

Universal tumor screening in a population with MSH6- and PMS2-associated Lynch syndrome

PurposeUniversal screening for Lynch syndrome (LS) on resected colorectal carcinomas (CRCs) and endometrial carcinomas (ECs) was implemented in Iceland in 2017 using immunohistochemistry (IHC) for mismatch repair (MMR) proteins. We examined the efficacy of the universal screening algorithm to detect LS and the diagnostic accuracy of MMR IHC by comparing results with a population-based genotype database.MethodsAll patients diagnosed with CRC or EC per the Icelandic Cancer Registry from 2017 to 2019 who had tumor MMR IHC performed were included. Pathology reports and patient charts were reviewed. MMR IHC stains were crossmatched with genotyping results obtained from the deCODE database.ResultsIHC staining was done on 404 patients with CRC and 74 patients with EC. A total of 61 (15.1%) patients with CRC and 15 (20.3%) patients with EC were MMR-deficient. MMR IHC had 88.9% sensitivity in identifying patients with LS and a positive predictive value of 10.7%. Only 50% of individuals were appropriately referred for genetic testing, leading to underdiagnosis of LS.ConclusionUniversal screening for LS using MMR protein IHC in CRC and EC accurately identified patients appropriate for genetic testing in a population with MSH6 and PMS2 LS predominance. Because of lack of referral to genetic counseling, only 50% of patients with LS were identified through the screening algorithm.     

Unrecognized norovirus infections in health care institutions and their clinical impact

Noroviruses (NoVs) have emerged as the leading cause of acute viral gastroenteritis (GE) in humans. Although diagnostic facilities have greatly improved, significant underdiagnosis of NoV in hospitals may still occur, thereby increasing clinical burden and nosocomial spread. We evaluated the underdiagnosis of sporadic NoV infections in a tertiary care hospital and estimated its clinical impact. From December 2008 until July 2009, fecal samples specifically referred for bacterial but not viral examination were retrospectively tested for NoV by real-time PCR. The clinical and virological data from patients with undiagnosed NoV infection (missed patients) were evaluated and compared with those from patients with recognized NoV. During the study period, 45 patients with undiagnosed NoV were detected, whereas 50 patients were regularly diagnosed. The missed NoV cases were more frequently adults than children (80% versus 46%; P < 0.001). The viral load levels did not differ between the diagnosed and missed patients, but missed patients more frequently presented without diarrhea (20% versus 4%; P < 0.07). The newly admitted missed NoV cases with GE underwent more diagnostic imaging (24% versus 4%; P < 0.01) and tended to be hospitalized longer. When missed-NoV patients were included, the number of nosocomial clusters doubled and missed patients were index cases in 5 of the 6 clusters. These data indicate that NoV infections are frequently missed despite routine laboratory testing and demonstrate that underdiagnosis of NoV patients is associated with costly abdominal imaging and nosocomial clustering. Awareness of NoV infection in adult patients and education about the importance of viral GE should be increased.     

Global study of primary immunodeficiency diseases (PI)—diagnosis,treatment, and economic impact: an updated report from the Jeffrey Modell Foundation

A large population of patients with recurring infections are undiagnosed or under diagnosed and Primary Immunodeficiency (PI) is more common than had been previously estimated. The results strongly indicate the measurable impact of Physician Education and Public Awareness in identifying patients with an underlying PI. The Jeffrey Modell Centers Network (JMCN) provides the infrastructure for referral, diagnosis and appropriate treatment. All disease classifications and identified defects increased significantly over the study period. Quality of Life for referred and diagnosed patients dramatically improved compared to undiagnosed patients. There is a substantial cost savings for diagnosed patients compared to undiagnosed, even if regular IgG is required. The SPIRIT^® Software successfully identified patients with PI in a large database and at three pilot sites. The Software was successfully tested for specificity and sensitivity.     

Clinical outcomes of a genomic screening program for actionable genetic conditions

PurposeThree genetic conditions—hereditary breast and ovarian cancer syndrome, Lynch syndrome, and familial hypercholesterolemia—have tier 1 evidence for interventions that reduce morbidity and mortality, prompting proposals to screen unselected populations for these conditions. We examined the impact of genomic screening on risk management and early detection in an unselected population.MethodsObservational study of electronic health records (EHR) among individuals in whom a pathogenic/likely pathogenic variant in a tier 1 gene was discovered through Geisinger’s MyCode project. EHR of all eligible participants was evaluated for a prior genetic diagnosis and, among participants without such a diagnosis, relevant personal/family history, postdisclosure clinical diagnoses, and postdisclosure risk management.ResultsEighty-seven percent of participants (305/351) did not have a prior genetic diagnosis of their tier 1 result. Of these, 65% had EHR evidence of relevant personal and/or family history of disease. Of 255 individuals eligible to have risk management, 70% (n = 179) had a recommended risk management procedure after results disclosure. Thirteen percent of participants (41/305) received a relevant clinical diagnosis after results disclosure.ConclusionGenomic screening programs can identify previously unrecognized individuals at increased risk of cancer and heart disease and facilitate risk management and early cancer detection.     

Designing for privacy management in hospitals: Understanding the gap between user activities and IT staff’s understandings

PurposeWe examined the role of privacy in collaborative clinical work and how it is understood by hospital IT staff. The purpose of our study was to identify the gaps between hospital IT staff members’ perceptions of how electronic health record (EHR) users’ protect the privacy of patient information and how users actually protect patients’ private information in their daily collaborative activities. Since the IT staff play an important role in implementing and maintaining the EHR, any gaps that exist between the IT staff’s perceptions of user work practices and the users’ actual work practices can result in a number of problems in the configuration, implementation, or customization of the EHR, which can lead to collaboration challenges, interrupted workflow, and privacy breaches.MethodsWe used qualitative data collection methods for this study. We conducted semi-structured interviews with 20 hospital IT staff members. We also conducted observations of EHR users in the in-patient units of the same hospital.ResultsWe identified gaps in IT staff’s understandings of users’ work activities, especially in regards to privacy-compromising workarounds that are used by users and why they are used.DiscussionWe discuss the reasons why this gap may exist between IT staff and users and ways to improve IT staff's understanding of why users perform certain privacy-compromising workarounds.ConclusionA hospital’s IT staff face a daunting task in ensuring users’ collaborative work practices are supported by the system while providing effective privacy mechanisms. In order to achieve both goals, the IT staff must have a clear understanding of their users’ practices. However, as this study highlights, there may be a mismatch between the IT staff’s understandings of how users protect patient privacy and how users actually protect privacy.     

Penetrance and outcomes at 1-year following return of actionable variants identified by genome sequencing

PurposeWe estimated penetrance of actionable genetic variants and assessed near-term outcomes following return of results (RoR).MethodsParticipants (n = 2,535) with hypercholesterolemia and/or colon polyps underwent targeted sequencing of 68 genes and 14 single-nucleotide variants. Penetrance was estimated based on presence of relevant traits in the electronic health record (EHR). Outcomes occurring within 1-year of RoR were ascertained by EHR review. Analyses were stratified by tier 1 and non–tier 1 disorders.ResultsActionable findings were present in 122 individuals and results were disclosed to 98. The average penetrance for tier 1 disorder variants (67%; n = 58 individuals) was higher than in non–tier 1 variants (46.5%; n = 58 individuals). After excluding 45 individuals (decedents, nonresponders, known genetic diagnoses, mosaicism), ≥1 outcomes were noted in 83% of 77 participants following RoR; 78% had a process outcome (referral to a specialist, new testing, surveillance initiated); 68% had an intermediate outcome (new test finding or diagnosis); 19% had a clinical outcome (therapy modified, risk reduction surgery). Risk reduction surgery occurred more often in participants with tier 1 than those with non–tier 1 variants.ConclusionRelevant phenotypic traits were observed in 57% whereas a clinical outcome occurred in 19% of participants with actionable genomic variants in the year following RoR.     

Closing the gap: Identification and management of familial hypercholesterolemia in an integrated healthcare delivery system

BackgroundFamilial hypercholesterolemia (FH) is an autosomal dominant genetic disorder that causes markedly elevated risk for early onset coronary artery disease. Despite availability of effective therapy, only 5–10% of affected individuals worldwide are diagnosed.ObjectiveTo develop and evaluate a novel approach for identifying and managing patients with FH in a large integrated health system with a diverse patient population, using inexpensive methods.MethodsUsing Make Early Diagnosis/Prevent Early Death (MEDPED) criteria, we created a method for identifying patients at high risk for FH within the Kaiser Permanente Northern California electronic medical record. This led to a pragmatic workflow for contacting patients, establishing a diagnosis in a dedicated FH clinic, and initiating management. We prospectively collected data on the first 100 patients to assess implementation effectiveness.ResultsNinety-three (93.0%, 95%CI: 86.1%–97.1%) of the first 100 evaluated patients were diagnosed with FH (median age = 38 years) of whom only 5% were previously recognized; 48% were taking no lipid-lowering therapy, and 7% had acute coronary symptoms. 82 underwent successful genetic testing of whom 55 (67.1%; 95%CI: 55.8%–77.1%) had a pathogenic mutation. Following clinic evaluation, 83 of 85 (97.6%) medication-eligible patients were prescribed combination lipid-lowering therapy. 20 family members in the healthcare system were diagnosed with FH through cascade testing.ConclusionsThis novel approach was effective for identifying and managing patients with undiagnosed FH. Care gaps in providing appropriate lipid-lowering therapy were successfully addressed. Further development and dissemination of integrated approaches to FH care are warranted.     

Interventions to improve delivery of cancer genetics services in the United States: A scoping review

PurposeInterventions that decrease barriers and improve clinical processes can increase patient access to guideline-recommended cancer genetics services. We sought to identify and describe interventions to improve patient receipt of guideline-recommended cancer genetics services in the United States.MethodsWe performed a comprehensive search in Ovid MEDLINE and Embase, Scopus, and Web of Science from January 1, 2000 to February 12, 2020. Eligible articles reported interventions to improve the identification, referral, genetic counseling (GC), and genetic testing (GT) of patients in the United States. We independently screened titles and abstracts and reviewed full-text articles. Data were synthesized by grouping articles by clinical process.ResultsOf 44 included articles, 17 targeted identification of eligible patients, 14 targeted referral, 15 targeted GC, and 16 targeted GT. Patient identification interventions included universal tumor testing and screening of medical/family history. Referral interventions included medical record system adaptations, standardizing processes, and provider notifications. GC interventions included supplemental patient education, integrated GC within oncology clinics, appointment coordination, and alternative service delivery models. One article directly targeted the GT process by implementing provider-coordinated testing.ConclusionThis scoping review identified and described interventions to improve US patients’ access to and receipt of guideline-recommended cancer genetics services.     

Using population-based data to predict the impact of introducing noninvasive prenatal diagnosis for Down syndrome

PurposeTo compare the number and types of chromosome abnormalities prenatally diagnosed and the number of invasive procedures between current prenatal testing pathways and a pathway where noninvasive prenatal diagnosis for Down syndrome replaces Down syndrome screening tests.MethodsNumbers and types of chromosome abnormalities for each referral category were extracted from prenatal diagnostic testing reports routinely collected in Victoria, Australia, in 2006 and 2007. These data were then applied to the proposed implementation strategy.ResultsIf noninvasive prenatal diagnosis for Down syndrome had replaced Down syndrome screening tests in 2006 and 2007, in Victoria, there would have been 25 (7%) additional Down syndrome diagnosed, 6896 (84%) fewer invasive procedures, and 231 (56%) non-Down syndrome chromosome abnormalities no longer detected. These include trisomy 13, trisomy 18, sex chromosome abnormalities, balanced and unbalanced rearrangements, polyploidy, and mosaic results.ConclusionsThe potential loss of information about chromosome abnormalities other than Down syndrome with noninvasive prenatal diagnosis compared with full karyotyping with traditional prenatal diagnosis should be considered when planning for the implementation of new technologies.