The studio authoring application starts with the creation of the patient. Authors can set the patient’s age, sex, appearance, vitals and personality. They then choose modules to build a patient case. Finally, the system walks authors through each module.

Module creation is designed for ease of use and walks the author through each step, very much like an on-line tax preparation program. Structured encounter modules are also available as an advanced option. Authors can also take an existing VSP and modify it or borrow assets from the sharing community.

Due to the fact that VSP encounters previously required an entire team to author them over a period of four to six months, there are very few VSPs in existence. With Standard Patient Studio, a single medical educator can create a working medical interview in twenty minutes and author a rich, multi-featured case with embedded assessment in one working day. This is made possible due to 1) guided authoring, 2) the unified medical taxonomy (UMT) and the ability of the authoring system to prepopulate ‘normal’ value medical responses and unique labs for each patient.

Assessment is built into authoring from the ground up. The author marks important parts of the medical taxonomy with assessment tags. Tags indicate the importance of eliciting a taxonomy item and what its relationship is to the diagnosis or alternative considerations. As the case gets tagged, the author creates a unique case taxonomy that will serve as the model for evaluating the learner’s case specific performance.

The hospital is the component that learners use to interact with virtual standardized patients. The hospital contains a number of patients, called VSPs.

VSP cases can be quite varied and may consist of one or more modules. Modules include the medical interview, structured encounters, physical exam, objective tests, general multimedia, anticipatory guidance, diagnosis & plan and assessment reporting. The hospital application stitches these modules together into a coherent whole which appears to the learner as a patient encounter.

Structured patient encounters involve choices that are presented to the learner. The virtual human character can be the VSP, a relative of the patient or the attending physician. Conversations involve three to five choices each turn. Conversations can branch and lead to vastly different outcomes. Structured encounters allow for mixed patient-learner initiative and provide for more provocative and varied experiences. Possible structured encounters include a patient interview, patient counseling, interaction with a third party character, or a reflective encounter with the attending physician. This approach has been successfully employed in ICT communications training projects like INOTS and ELITE.

The medical interview is a natural language verbal encounter between the learner and the virtual standardized patient (VSP). The learner initiates questions by typing them or speaking to the computer. The VSP responds and returns control to the learner. Beginners can benefit from pedagogical guidance features for their first few patients. After familiarization, learners are expected to display good bedside manner, discover patient concerns by drawing out a narrative and asking specific questions to establish a diagnosis. This medical interview is the default interaction for Standard Patient Hospital encounters.

VSPs themselves have a variety of baseline personalities and some are more loquacious, defensive or articulate than others. The patients also judge the learner in real-time through use of the Inference-RTS™ assessment engine. They may say more to friendlier doctors or disclose sensitive information to doctors they view to be more competent and trustworthy. Patients are also capable of deception.

A wide range of tests are available. This includes the physical examination, laboratory tests, radiological tests, physiological tests and other standard assessments. It is up to the learner to select the most appropriate actions within this standardized repertoire based upon the patient complaint and the interview. Results appear as they typically would in a hospital computer system. VSP authors can include photos, sound and video into any result and query the learner for their interpretation.

A patient encounter is not complete without intelligent guidance on performance. The assessment system will offer a global assessment of performance and detailed assessments for each modular component of the VSP encounter. The system will determine the learner’s level of diagnostic maturity and provide feedback for improvement within the learner’s proximal level of development.

The Teacher’s Lounge is a light learning management system intended for daily use by medical educators. This system can be used to build courses out of several cases or assign a case for testing purposes and can enroll students in the teacher’s courses. It includes access to individual student case assessment reports, overall student metrics and a class wide performance dashboard.

The three audience areas for Standard Patient include medical learners, mature clinicians & medical educators. For medical learners, the optimal audience ranges from third year of medical school through residency. The focus is to more rapidly mature their interviewing skills and in the future to provide a controlled case curriculum as a basis for establishing competency in a clinical area.

For mature clinicians, Standard Patient can be used for case-specific competency assessment, refresher training or continuing medical education. Medical educators can author cases, take advantage of the sharing community to obtain VSP patients and assets, and review student assessments.

Three ICT technologies power Standard Patient; SimCoach, the Unified Medical Taxonomy (UMT) and the Inference-RTS Assessment Engine. SimCoach is ICT’s web-based virtual human platform that allows for real-time natural language virtual human interactions through a web browser or tablet. The Unified Medical Taxonomy is a common data model that represents a human patient and contains over 1,200 medical items for each patient. The Inference-RTS Assessment Engine is a real-time artificial intelligence system that continuously evaluates learner performance during the medical encounter and aggregates feedback into useful guidance for improvement.

When available in 2015, Standard Patient will be free for anyone to use. The authoring system will also be free. There is a catch – this is a sharing project. Any case that is authored is automatically accessible to all medical educators in the community. The intent is to create a critical mass of high quality patients across a range of specialties.

The USC Standard Patient is a new University of Southern California Institute for Creative Technologies research project. It seeks to improve clinic-based medical encounter simulation with the goal to create engaging virtual standardized patient (VSP) encounters, enable objective and meaningful assessment of learner interview performance and mature physician interviewing & diagnostic skills.

Standard Patient is a modular system that includes several complete applications that work together. There is an encounter player, an authoring application, an on-line community and a learning management system.

A Virtual Standardized Patient is a virtual human that speaks, expresses him or herself nonverbally and understands your questions in normal clinical language. Standard Patient Hospital features more than fifty different virtual humans that vary by age, ethnicity and body type. VSP encounters closely simulate office-based patient visits where the clinician must obtain a chief complaint, elicit a narrative and follow up with specific questions to establish a diagnosis and rule out other conditions within a working differential.

Typical ‘Virtual Patients’ are usually text-based cases that sometimes have multimedia elements and different possible outcomes in their progress based upon learner choices as the case unfolds. Virtual patients are typically used to assess medical decision making. Sometimes they are used to determine if the learner recognizes a diagnosis or has specific knowledge. Although there is a medical decision making component to VSPs, their richest benefits are concerned with case-specific diagnostic skills and strategies. VSPs are substantially different in design and use from what is typically referred to as Virtual Patients.

Standard Patient can assess a variety of skills in this area. For entry level learners, the system assesses general interviewing skills such as interpersonal skills, open ended questioning, solicitation of a patient narrative, choice of follow up questions and demonstration of empathy and respect for the patient. For more advanced learners, interview strategy is more diagnosis-specific and involves elicitation of an illness script and pertinent negative follow up questions to narrow the differential. The learner’s diagnostic strategy can be determined as can the learner’s efficiency and overall diagnostic skills maturity.

Since the VSP encounter is intended to resemble a single clinic visit, there are no opportunities for a VSP to die on an operating table or to miraculously improve from the learner’s valiant efforts. Medical decision making assessment is based upon the learner’s seUSC Standard Patientlection of diagnostic tests, identification of main and supporting diagnoses and selection of appropriate medications, procedures and therapies. VSP cases may also contain a dynamic interview with Hippocrates, the USC Standard Patient Attending Physician. The attending physician encounter can be used to explore reasons behind the learner’s decision making or their patient management strategy.

The American College of Graduate Medical Education (ACGME) competency milestone concept has significantly shaped the design of Standard Patient. Specifically, elements of the interview assessment system have been inspired by the American Board of Pediatrics (ABP) milestone project. ABP milestones uniquely scale from entry to medical school through continuing medical education. The areas addressed by Standard Patient include 1) Patient Care, 2) Medical Knowledge and 3) Interpersonal & Communications Skills. The assessment system will provide feedback in context of milestones when possible.