Group items tagged

Our payer mix included 7.2% of payments from Medicaid (exclusively through Medicaid health maintenance organizations), 21.5% from Medicare (of which 14.0% were fee-for-service and 7.5% capitated), 64.7% from commercial insurers (34.5% fee-for-service and 30.2% capitated), and 6.5% from pay-for-performance programs.

Throughout 2008, our physicians provided 118.5 scheduled visit-hours per week, ranging from 15 to 31 weekly hours each. We regard this schedule as equivalent to the work of four full-time physicians, with physicians typically working 50 to 60 hours per week. Our staff included four medical assistants, five front-desk staff, one business manager, one billing manager, one health educator (hired midyear), and two full-time clerical staff. Our staffing ratio was approximately 3.5 full-time support staff per full-time physician. We had no nurses or midlevel practitioners.

We use an electronic health record, which we adopted in July 20043 and use exclusively to store, retrieve, and manage clinical information. Our electronic system came with 24 "document types" that function like tabs in a paper chart to organize documents, dividing clinical information into categories such as "office visit," "phone note," "lab report," and "imaging." Since all data about patients is stored in the electronic record (either as structured data or as scanned PDFs) and each document is signed electronically by a physician, we are able to measure accurately the volume of documents, which serve as proxies for clinical activities, in a given time period.

The volume and types of documents that we receive, process, and create are listed in Table 1

Telephone calls that were determined to be of sufficient clinical import to engage a physician averaged 23.7 per physician per day, with 79.7% of such calls handled directly by physicians.

Of these calls, 35.7% were for an acute problem, 26.0% were for administrative purposes

Physicians averaged 16.8 e-mails per day. Of these electronic communications, 59.3% were for the interpretation of test results, 21.7% were for response to patients (either initiated by patients through the practice's interactive Web site or as part of an e-mail dialogue with patients), 9.3% were for administrative problems, 5.0% were for acute problems, 2.8% were for proactive outreach to patients, and 1.9% were for discussions with consultants.

Each physician reviewed 19.5 laboratory reports per day, including those ordered through our office (which are delivered to us through an electronic interface and are automatically posted to the database of the electronic health record as numerical values) and those ordered outside our office (which enter our chart as scanned PDFs and are not posted as numerical values). The work cycle of responding to a laboratory result includes interpretation by telephone, letter, or e-mail. (Our office sent 12,541 letters communicating test results, about a third of which were sent by e-mail.) For noninterfaced laboratories, we must decide which values need to be entered manually into the electronic health record by a staff person; the values of scanned results cannot be graphed or searched without this step. Laboratory results frequently trigger a review or adjustment of a medication, which requires access to accurate, current medication lists with doses.

Each physician reviewed 11.1 imaging reports per day, which usually required communication with patients for interpretation. Such review may require updating problem lists (e.g., a new diagnosis of a pulmonary nodule) or further referral (e.g., fine-needle aspiration for a cold thyroid nodule), which generates additional work, since results and recommendations are communicated to patients and consultants.

Each physician reviewed 13.9 consultation reports per day. Such reports from specialists may require adjustments to a medication list (if a specialist added or changed a medication), changes to a problem list, or a call or e-mail to a patient to explain or reinforce a specialist's recommendation. Some consultation or diagnostic reports relate to standard quality metrics (e.g., eye examinations for patients with diabetes) and need to be recorded in a different manner to support ongoing quality reporting and improvement.5

Before our practice had an electronic health record, we employed a registered nurse. After the implementation of the electronic health record system, much of the work that the nurse performed could be done by staff who did not have nursing skills, and by 2008, we no longer employed a registered nurse. However, on the basis of the analysis described here, we have hired a registered nurse to do "information triage" of incoming laboratory reports, telephone calls, and consultation notes — a completely different job description than what we had before.

Our practice is participating in a multipayer Patient Centered Medical Home demonstration project7 (which allowed us to hire our health educator). This project is overseen by the Pennsylvania governor's office and funded by the three largest commercial insurers and all three Medicaid insurers in our region

This process can be very human-centered, as in a journal club, or highly technical, using computer programs and information techniques such as data mining.

Level III: Opinions of respected authorities, based on clinical experience, descriptive studies, or reports of expert committees.

Despite the differences between systems, the purposes are the same: to guide users of clinical research information about which studies are likely to be most valid. However, the individual studies still require careful critical appraisal.

The U.S. Preventive Services Task Force uses:[9] Level A: Good scientific evidence suggests that the benefits of the clinical service substantially outweighs the potential risks. Clinicians should discuss the service with eligible patients. Level B: At least fair scientific evidence suggests that the benefits of the clinical service outweighs the potential risks. Clinicians should discuss the service with eligible patients. Level C: At least fair scientific evidence suggests that there are benefits provided by the clinical service, but the balance between benefits and risks are too close for making general recommendations. Clinicians need not offer it unless there are individual considerations. Level D: At least fair scientific evidence suggests that the risks of the clinical service outweighs potential benefits. Clinicians should not routinely offer the service to asymptomatic patients. Level I: Scientific evidence is lacking, of poor quality, or conflicting, such that the risk versus benefit balance cannot be assessed. Clinicians should help patients understand the uncertainty surrounding the clinical service.

AUC-ROC The area under the receiver operating characteristic curve (AUC-ROC) reflects the relationship between sensitivity and specificity for a given test. High-quality tests will have an AUC-ROC approaching 1, and high-quality publications about clinical tests will provide information about the AUC-ROC. Cutoff values for positive and negative tests can influence specificity and sensitivity, but they do not affect AUC-ROC.

In 2007 tragedy ended their long partnership. Mr. Gray set out from San Francisco on a solo trip on his 40-foot sailboat and did not return.

A couple of years after Mr. Szalay joined the project, a colleague introduced him to Jim Gray, who was a kind of rock star himself—in the computer-science world. Wired magazine once wrote that the programmer's work had made possible ATM machines, electronic tickets, and other wonders of modern life. When Mr. Szalay met him, Mr. Gray was a technical fellow at Microsoft Research and was looking for enormous sets of numbers to place in the databases he was designing.

in 1992 came the project that would change his career. Johns Hopkins joined the Sloan Digital Sky Survey project, a computerized snapshot of the heavens.

The scientists, along with tech-industry leaders whom Mr. Gray had mentored in the past, offered to help the Coast Guard search the open sea using any technology they could think of. Google executives and others helped provide fresh satellite images of the area. And an official at Amazon used the company's servers to send those satellite images to volunteers—more than 12,000 of them stepped forward—who scanned them for any sign of the lost researcher.

But Jim Gray was never found. Some of the techniques that the astronomer learned from the search effort, though, have now been incorporated into a Web site that invites anyone to help categorize images from the Sloan Digital Sky Survey.

The number of volunteers surprised the organizers. "The server caught fire a couple of hours after we opened it" in July 2007, he said, burning out from overuse. More than 270,000 people have signed up to classify galaxies so far.

Gene Wikis

It started under the name of GenMAPP, or Gene Map Annotator and Pathway Profiler. Participation rates were low at first because researchers had little incentive to format their findings and add them to the project. Tenure decisions are made by the number of articles published, not the amount of helpful material placed online. "The academic system is not set up to reward the sharing of the most usable aspects of the data," said Alexander Pico, bioinformatics group leader and software engineer at the Gladstone Institute of Cardiovascular Disease. In 2007, Mr. Pico, a developer for GenMAPP, and his colleagues added an easy-to-edit Wiki to the project (making it less time-consuming to participate) and allowed researchers to mark their gene pathways as private until they had published their findings in academic journals (alleviating concerns that they would be pre-empting their published research). Since then, participation has grown quickly, in part because more researchers—and even some pharmaceutical companies—are realizing that genetic information is truly useful only when aggregated.