A novel electronic consult (e-consult) system for a pathology and laboratory medicine service (PLMS) was implemented in 2015 at a high-complexity Veterans Administration health care facility. Consults were previously made through direct provider communication without documentation in the medical record.
To evaluate the utilization trends of the laboratory e-consult system at the Department of Veterans Affairs Connecticut facility during the first 2 years since inception.
E-consultation involves pathology and laboratory medicine resident review followed by attending pathologist review and cosignature. E-consults to the pathology and laboratory medicine service from 2015 to 2017 were reviewed to record type of consult, requesting department, patient location, and turnaround time.
The pathology and laboratory medicine service received 351 e-consults from 2015 to 2017. The volume varied by subsection: hematology and coagulation (215 of 351; 61%), chemistry (109 of 351; 31%), blood bank (19 of 351; 6%), and microbiology/virology (8 of 351; 2%). Hematology and coagulation consults were entirely for peripheral blood smear review (215 of 215; 100%). Chemistry consults were placed for toxicology/drugs of abuse (81 of 109; 74%), test utilization (17 of 109; 16%), or nontoxicology (11 of 109; 10%). Three services placed the majority of consults: primary care (279 of 351; 80%), hematology/oncology (39 of 351; 11%), and psychiatry (27 of 351; 8%). The median turnaround time for completion of e-consults was 1.2 days. Since e-consult implementation, the mean number of consults increased from 8.6/mo in 2015 to 18.1/mo in 2017, peaking in the last quarter of analysis in 2017 with a mean of 25.3 consults/mo.
This novel e-consult system improved accessibility to and documentation of answers to laboratory questions and increased the visibility of the pathology and laboratory medicine service. Future goals include development of outcomes-based measures to better assess the clinical impact of e-consults.
Electronic consult (e-consult) systems have emerged as a potential solution to increasing demands within health care systems. E-consults are an electronic method for primary providers to request input from specialty services regarding patient management.
Many benefits of e-consult systems have been reported such as improved access to, and decreased time to see, consultants for specialty care and improved overall workflow efficiency. Importantly, e-consults occur within the electronic health record (EHR), allowing consultants to rapidly access patient information and provide formal documentation of recommendations within the EHR. E-consult systems reduce patient wait times for specialty care,1 with the time to completion of e-consults typically less than 3 days. Reviews demonstrate up to a 51% decrease in the number of specialist visits associated with use of e-consultation.2 Additionally, studies examining primary care physician satisfaction with e-consult systems report good satisfaction, ranging from 70% to 95%.3–11 For example, 71% of primary care physicians at San Francisco General Hospital felt that e-consultation systems improved overall clinical care.9 Specialist satisfaction with e-consultation has been mixed, likely reflecting concerns about medico-legal liability and the need for protected time to complete the e-consults.12
To date, several academic medical centers have established e-consult systems. Vimalananda et al12 performed a systematic review of the e-consult literature between 1990 and 2014 and found that 3 health care systems, including San Francisco General Hospital, Mayo Clinic, and the Department of Veterans Affairs (VA), accounted for the majority of publications related to experience with e-consult systems. The Veterans Health Administration's Specialty Care Transformation Program Office began implementing e-consults in 2011.13 E-consults are now used by more than 50 medical specialties in 152 medical centers and more than 800 community-based outpatient clinics within the Veterans Health Administration, but to date there are no reports on such consults being used by pathology or laboratory medicine providers.
E-consults have the potential to increase provider awareness of the spectrum of activities and services that laboratory medicine provides. Many clinical decisions are based on laboratory tests results; therefore, collaboration between laboratory medicine physicians and bedside clinicians is critical for patient care. A recent survey conducted by the Center for Disease Control and Prevention's Clinical Laboratory Integration Into Healthcare Collaborative of general internal medicine and family medicine physicians suggests there is room for improvement in communication between laboratory medicine physicians and clinicians. In 14.7% of patient encounters, physicians were uncertain of which tests to order, and in 8.3% of patient encounters, physicians had difficulty interpreting test results.14 Laboratory medicine physicians also possess high levels of expertise regarding laboratory tests, their utilization, and their interpretation, but physicians were not likely to consult laboratory medicine to address such questions. There are several potential reasons for this lack of communication between physicians and laboratory medicine physicians. For example, the survey found that bedside physicians were unsure of the right person to call in the laboratory. This survey thus highlights a need for implementation of e-consult systems within laboratory medicine to help address barriers in communication between laboratory medicine physicians and clinicians. In addition, the e-consult system has the potential to allow for legal billing for services rendered by laboratory medicine providers. Reports of successful implementation of e-consult systems are well documented in a variety of medical specialties, including hematology, endocrinology, infectious disease, and cardiology, among others4,15–17 ; however, reports of use of e-consults within laboratory medicine are scarce.18
With the goal of addressing these challenges and increasing provider access to laboratory medicine services, the pathology and laboratory medicine service (PLMS) at VA Connecticut (VA CT) implemented an e-consult system in 2015. Prior to implementation of the e-consult system, consults were typically performed through informal direct provider communication without documentation in the EHR. The aim of this study was to describe the establishment of the e-consult system and evaluate trends in utilization of e-consults by medical providers during the first 2 years since its inception, from 2015 to 2017.
This study was conducted at VA CT, a high-complexity Veterans Administration Healthcare facility composed of 2 major care facilities in the state (ie, the West Haven, Connecticut, campus and the Newington, Connecticut, campus). Combined, these facilities accounted for 194 inpatient beds, 148 147 outpatient visits, and 2 438 897 laboratory tests performed in 2017. In 2019, the year in which we performed our lookback on e-consult utilization, laboratory volumes at VA CT varied by service as follows: 107 082 hematology and coagulation tests, 1 345 932 chemistry tests, 10 610 blood bank tests, and 49 589 microbiology and virology tests. In addition, VA CT is a major academic affiliate of the Yale School of Medicine (New Haven, Connecticut) and the Laboratory Medicine service at VA CT is covered 24/7 by a resident from the Yale Pathology Residency Program.
Figure 1 summarizes the basic workflow for handling e-consult requests received by PLMS at VA CT. E-consults to PLMS are directly entered into the VA EHR by requesting providers as consults for “clinical pathology and lab e-consult.” Figure 2 shows an example of an e-consult request from the EHR using an artificial “test patient.” The request contains the patient location (inpatient versus outpatient), requesting provider and service department, urgency of consult, and reason for request. The requesting provider uses free text to enter the consult question into the reason for request field. If a provider has a question regarding a particular specimen or result, or has a specific request, then the relevant information such as specimen type, date, and information of interest are included by the provider as free text in the ‘reason for request' field.
Residents on the VA laboratory medicine service receive printout paper copies of new e-consults received by the PLMS as part of their daily work. An example printout of an e-consult using a test patient is shown in Figure 3. The working copy field is where results of e-consults appear once completed by PLMS. As e-consults are received, residents review the patient records, including provider notes, patient medications, and pertinent laboratory data, and generate consult notes within the EHR with initial impressions and recommendations. A document containing detailed step-by-step instructions with example screenshots on how to enter and sign e-consult notes in the EHR is also available to residents on an internal shared drive. In addition, residents have access to pertinent references and instructions for handling common problems (eg, in toxicology or drugs of abuse test interpretation). Templates for other common consults (eg, peripheral smear reviews) were also developed to standardize the process.
After a preliminary note is completed by the resident in the EHR, the e-consult is evaluated at daily sign-out with the on-service laboratory medicine attending pathologist, who rereviews the case and cosigns the resident's note. Preliminary results entered by the resident are not visible to the ordering provider until the note is signed by the laboratory medicine attending pathologist. Once the note is signed by the laboratory medicine attending pathologist, the note appears in the note section of the EHR and the ordering provider receives an inbox alert that the e-consult is complete. Potential outcomes of the e-consult include request for more information, clarification of consult question, or final recommendations (Figure 1). In the case of an urgent/stat laboratory question, the residents can be reached by telephone as they provide 24/7 on-call services with attending pathologist backup. Clinicians can obtain the phone number of the on-call PLMS resident through the laboratory/blood bank technical staff or the VA operator.
Total e-consults placed to PLMS between 2015 and 2017 were extracted from the EHR for the 2 VA CT care facilities. E-consults were advertised to providers as offering consultative requests specifically within the domain of laboratory medicine, to include consults in one or more of the 4 following areas: hematology and coagulation, chemistry, blood bank and transfusion medicine, and microbiology and virology. Consults were subsequently categorized for this study into 1 of 4 laboratory domains by reading the free text entered by ordering providers in the reason for request field of the e-consult. By design, e-consults are defaulted to moderate priority in the EHR, with urgent requests made via direct contact of the on-call clinical pathology resident/attending pathologist. For simplicity, molecular diagnostic e-consults were grouped with the core laboratory that most closely relates to the specific molecular question (eg, hepatitis C genotyping would be grouped with microbiology/virology). Consults made to the laboratory medicine service that were outside of the core laboratory medicine domains listed above were recorded for the purpose of identifying the number of inappropriate consults made to our service; in practice, such consults were either forwarded to the appropriate service or cancelled after discussion with the requesting provider. Information collected from each e-consult for this study included the type of e-consult question, date of request, turnaround time (time from placement of e-consult request by ordering provider in EHR to signature of PLMS note by laboratory medicine attending pathologist), requesting facility, department specialty requesting e-consult, and patient location (inpatient versus outpatient). Institutional review board approval was obtained for this study.
A total of 351 e-consults were placed to PLMS from 2015 to 2017. Nine services placed e-consults, with the majority from 3 services: primary care (279 of 351; 80%), hematology/oncology (39 of 351; 11%), and psychiatry (27 of 351; 8%). Additional services that placed e-consult requests included dermatology, gastroenterology, ophthalmology, pulmonology, rheumatology, and social work. The majority of the e-consults were requested in the outpatient setting (309 of 351; 88%). There were 21 inappropriate consults that were outside the scope of laboratory medicine, which were not included in the total 351 reported earlier. These inappropriate e-consults were all questions related to anatomical pathology (eg, interpretation of a biopsy specimen) and often represented duplication of information already submitted to the anatomical pathology division through a separate mechanism. These consults were canceled with a note in the EHR on how to follow up with the anatomical pathology division.
The volume of e-consults varied by 4 main laboratory subsections, as shown in Figure 4. Hematology and coagulation questions (215 of 351; 61%) constituted the majority of e-consults. Notably, all of the consults within the hematology and coagulation category were for peripheral blood smear interpretation (215 of 215; 100%). Chemistry e-consults constituted the second largest group of requests (109 of 351; 31%). These e-consults were further broken down into 3 subcategories: toxicology (81 of 109; 74%), test utilization (17 of 109; 16%), and other/nontoxicology (11 of 109; 10%), as shown in Figure 5. Toxicology consults largely consisted of interpretation of urine toxicology results in the context of the patient's prescribed medications. As an example, a clinician requested a toxicology e-consult to determine if a negative urine drug screen for benzodiazepines was consistent with regular use of clonazepam (1 mg taken twice a day). Test utilization consults included questions regarding selection of the proper laboratory test and how to place orders for a test within the EHR. Lastly, nontoxicology consults included questions such as the significance of icterus in a specimen, the accuracy of hemoglobin A1c assay for a patient with sickle cell disease, and the potential interference of intravenous immune globulin on tests such as hemoglobin A1c and prostate-specific antigen. Overall, there were no consults related to chemical test report interpretation, nor were there any consults regarding quality assurance activities in the chemistry laboratory.
Within the blood bank and transfusion medicine category (19 of 351; 6%), the majority of the e-consults were for immunohematology (15 of 19; 79%). The remaining blood bank e-consults were for evaluation of platelet clumping, appropriateness of platelet transfusion, and questions related to directed donation of blood products.
Only rare e-consults were submitted regarding microbiology and virology questions (8 of 351; 2%). Examples of consult questions in this domain included interpretation of laboratory results such as hepatitis C viral load, Borrelia burgdorferi serology, and Treponema pallidum serology and capabilities of testing for various infections/microorganisms on different specimen sources.
Turnaround time, defined as the time from placement of an e-consult request in the EHR by an ordering provider to the time of signature of the PLMS note by the laboratory medicine attending pathologist, was recorded for each e-consult. The median turnaround time for all consults was 1.2 days. Consults placed to blood bank had the shortest median turnaround time (0.9 day) followed by hematology and coagulation (1.1 days), chemistry (1.2 days), and microbiology and virology (1.8 days).
Throughout the study period, the number of e-consults received by PLMS continued to increase, as shown in Figure 6. In 2015, a mean of 8.6 e-consults per month were received, versus 18.1 in 2017. In the last quarter of 2017, a mean of 25.3 e-consults per month were received. Overall, these consults overwhelmingly represented first-time consults in unique patients (ie, those without a preexisting laboratory e-consult in the EHR), with duplicate requests of the same consult question on the same patient representing less than 1% of total e-consults.
Our novel PLMS e-consult system improved accessibility to and documentation of answers to complex laboratory questions. The total number of PLMS e-consults increased significantly during the study period, from 8.6 per month in 2015 to 18.1 per month in 2017. This is likely because of increased provider awareness and satisfaction with the e-consult system. Although physicians' satisfaction with the e-consult system was not measured in this study, a follow-up addendum by the physician for further clarification was issued on fewer than 5 e-consults. Based on these data, we infer that the e-consult questions were adequately answered. A future goal will be to directly survey providers to determine satisfaction and areas to improve the PLMS consultative e-consult service.
The overwhelming majority of e-consults placed to PLMS were for hematology consults, specifically for peripheral blood smear interpretation (215 of 351; 61%). The PLMS has a joint hematology rounds weekly conference where bone marrow biopsies, lymph nodes, and peripheral smears from the hematology service are reviewed. In addition, any test utilization, molecular hematopathology, or coagulation questions from clinicians are generally answered through informal discussions during the course of this conference, and these hematology questions are not captured as e-consults. This factor likely accounts for why the entirety of hematology e-consults consisted of peripheral smear review requests. On the other hand, there is no mechanism for nonhematology providers to obtain peripheral smear interpretation by a physician, another factor we believe contributes to the heavy use of the e-consult mechanism for smear review by nonhematology providers. Notably, our facility also hosts a weekly interdisciplinary microbiology rounds, one that is very diverse in nature discussing broad-based laboratory testing questions related to infectious disease; we speculate that this is one possible explanation for the lack of microbiology questions via the e-consult mechanism. Ultimately, future work (eg, surveys) will be required to better establish the patterns of use we have observed.
Within the chemistry category, the majority of e-consults were for interpretation of toxicology results (81 of 109; 74%). The large number of toxicology interpretation requests from providers highlights the laboratory medicine physician's useful role in complex interpretations for drug of abuse testing. On the other hand, although the test utilization subcategory comprised only 16% (17 of 109) of the total chemistry consults, we anticipate that requests in this domain will continue to increase as our laboratory test menu grows and the number of tests capable of being sent out to reference laboratories grows in parallel.
Blood bank and transfusion medicine e-consults consisted of 6% (19 of 351) of the total e-consults. One likely reason for the low number of consults is that providers are accustomed to reaching out directly to the blood bank director or pathology resident on service, particularly for high acuity concerns, thereby bypassing the e-consult system. Questions related to the blood bank service are often time sensitive, and immediate, direct phone conversations are often needed for timely patient care.
The findings from this study demonstrate that e-consults provide a bridge between clinicians and laboratory medicine physicians to provide patients with the best care. In addition, the e-consult system is a vital part of the pathology resident training program and offers trainees diverse experiences in acting as consultants. Unique to this consult mechanism, and in contrast to other forms of consultation that have been published in the domain of clinical pathology,18 is that the formal consult from providers outside of laboratory medicine and the written note also allow for potential patient billing and assignment of Current Procedural Terminology and International Classification of Diseases codes for each encounter. Thus, in addition to impacting patient care and providing a novel means for resident/trainee interaction with clinical teams, e-consults can be a highly useful means to demonstrate an alternative form of clinical pathologist value beyond typical duties. Our VA hospital does not directly bill patients for laboratory services; therefore, billing data were not included in this study. However, because relative value unit data are captured, the most commonly used Current Procedural Terminology codes for laboratory medicine consultation in descending order of use are summarized in the Table.
The e-consult system is not without challenges, and potential for improvement remains. For example, within the domain of hematology, there is opportunity for growth to provide feedback in other categories, such as coagulation, hemolysis laboratory interpretation, and body fluid analysis. Similar to data published elsewhere, test utilization consults, which included questions regarding selection of proper laboratory tests, constituted a substantial portion of e-consults, but there is also room for growth in this aspect of our practice. Moreover, microbiology and virology e-consults could be captured in the future by converting the informal discussions that occur at weekly microbiology plate rounds into e-consults placed in the EHR and/or encouraging additional consults via this platform for questions that arise during weekly rounds. In addition, informal consult questions received through telephone calls or during meetings that are not currently captured in the e-consult system continue to occur; more efforts could be made to request that clinicians convert such questions into formal consults in order to better document responses and avoid potential pitfalls of the undocumented curbside consult.
Finally, it is important to note that some previous studies of e-consult systems have focused on provider perceptions of such consult mechanisms. Although this was not performed in our analysis, we are planning future studies of our consult system with additional endpoints such as physicians' reaction to, or satisfaction with, the e-consult system. Such analyses are clearly needed to demonstrate not only that an e-consult service can be successful and diverse, but also that it is meeting the needs of ordering providers. Another major future goal of ours is the development of outcomes-based measures to more fully assess the clinical impact of our e-consult mechanism, including determining the total relative value units generated from the e-consult system, costs saved by this approach, and whether any safety benefits are realized with laboratory-based e-consults.
In summary, our experience implementing a laboratory e-consult system demonstrated that providers from different specialties used the system to address a variety of questions across the spectrum of clinical pathology. We believe the increasing trend in monthly e-consult requests demonstrates provider satisfaction and comfort level with the e-consult system and a desire for increased pathologist participation with patient care. We expect the number of e-consults to continue to rise and look toward generating outcomes-based analyses to further demonstrate the positive clinical impact of our consult approach.
We thank Wendy Strollo, MT, and Ankur Bhargava, MD, for their contributions to this study.
The authors have no relevant financial interest in the products or companies described in this article.