Final Group Presentation, Summary Document162389445480246740.weebly.com/uploads/1/4/5/9/... · Presentation Date: March 1, 2010 Page: 1 of 15 Final Group Presentation, Summary Document

402 Intro to Clinical Thinking, Winter 2010

Final Group Presentation, Summary Document

Group 1: Michael Aguilar, Hugo Fernandes, Charles Kitzman, Biljana Spasic

Presentation Date: March 1, 2010 Page: 1 of 15


The Introduction to Clinical Thinking class exposed us to the concepts,

advantages and issues related to the use of information technology in improving the

clinical process. In our Final Group Presentation Group 1 used a real life post mortem

case and applied the concepts taught in the class and investigated improvements enabled

by new technology and tools. To go through the clinical process we used the SOAP

approach which consists of the following steps:

1. Subjective in which medical history, history of the present illness and chief

complaint are collected and recorded.

2. Objective where the physical exam is performed and vitals are taken and

recorded.

3. Assessment to work diagnoses.

4. Plan which covers labs / diagnostic ordering and instructions for the patient.

Clinical case





Subjective

Our group used an example of an EMR system to collect, store and share the

medical history, history of the present illness and chief complaint. Having this

information available in one place was a visible improvement to the paper charts.

Sample EMR Interface

The advantages of an electronic medical record documentation and encounter

management include those listed below:

• Capture chief complaint / left side numbness, would have been more evident

• Enable retrieval of relevant data from previous visits, including stored diagnostic

tests, diagnostic procedures and labs

• Create physician-specific review of systems or chief complaint lists

• Categorize clinical data for problem management

• Drill down into summaries for all patient problems, history, encounters,

medications, orders and procedures

• Efficient management of problem list





Objective

Vitals

In the paper world, vitals are largely a single encounter concern as comparison

over time would require the maintenance of a grid style form that would require constant

upkeep and in many clinical settings, this isn’t the case. A two-fold advantage an EHR

provides in the area of vitals is the digital equipment capture of measurements and the

ability to grid and graph the results. The first reduces the human mistake factor though

clinical organizations must be diligent in making sure the equipment is properly

calibrated and that it receive routine maintenance or the supposed benefit can become an

enormous liability. Suppose that a good clinical staffer makes one mistake recording

vitals for every 25 patients. Then consider the number of inaccurate values if the

electronic equipment goes without maintenance.

The second benefit has obvious merits not only to the clinician to get a rapid and

accurate historical view of the patient’s vital statistic history, but it can also be useful in

engaging the patient in their own health trends whether they be positive or negative. A

patient trying to lose weight as part of their treatment plan would take comfort in a visual

representation of this historical data. A patient on a secondary medication to help reduce

blood pressure can compare the effectiveness over time of the newer medication to the

older one. As clinical firms move forward with EHR implementation, the personal

attention level that patients receive from clinicians will wane, sometimes quite a bit. The

use of visual tools such as graphs are good methods by which to engage and include the

patient when traditional personal “face-time” processes are reduced.

See images on the next page for example.





Sample images





Physical Exam

By the time a provider begins a physical exam, the Histories, Chief complaint,

HPI, and Review of systems will have already guided the decision process enough to

determine what areas will be examined. Their education and training on how different

areas of the body can be affected by others and symptomatic pathologies determine

“where to look”. An example would be that a patient presenting with an earache might

not warrant a genitourinary exam.

Physical exams and the info they supply are a crucial part of the encounter as they

can validate the decision-making process regarding lab ordering, diagnostic ordering,

treatment options, etc. As it relates to decision support, if an exam template is setup

properly, the discreet answer to common exam elements can potentially be fed into a

computer based decision support tools like DXplain or Isabel and if properly integrated,

assist the clinical teams in determining which orders give the greatest statistical

likelihood for assessment validation. It’s easy to imagine a short time from now having

EHR’s with fully integrated decision support mechanisms. In the following image, the

arrows next to the areas on general concern bode well for positive or negative values to

be fed, via trigger, to a decision support tool. This data in conjunction with the discrete

data captured at intake can further refine the technical decision making ability.





Assessment

Assessment involves using the patient history and physician examination to

identify the clinical diagnosis for the chief complaint. Assessment also involves creating

or updating the patient problem list. The clinical diagnosis for the chief complaint is

identified by a provider developing diagnostic possibilities for the complaint and then

testing each diagnostic possibility in order to confirm or rule out that possibility. This

process of developing and evaluating diagnostic possibilities can occur in multiple areas

of the clinical encounter. For example the process can occur during the taking of the

patient history and during the physical examination. Imaging, lab, and other tests can also

be used by the provider to rule out or confirm a clinical diagnosis.

Clinical diagnostic decision support tools could help providers in this area to

identify the clinical diagnosis. In many patient encounters a CDDS tool will not be

needed, because the provider can identify the clinical diagnosis by using data gathered

during the taking of the patient history and physical examination. There are however a

number of reasons why a provider may benefit from using a clinical diagnostic decision

tool. A provider may have a lack of knowledge about the causes for certain symptoms. A

provider may have identified a few diagnosis possibilities, but the provider wants to use

the tool to identify diagnostic possibilities that the provider may not have thought of.

There are a number of tools that can help providers to identify a clinical diagnosis. Two

tools are Isabel and Dxplain.

Isabel

Isabel is a Web-Based Clinical Diagnosis Support System developed by Isabel

Healthcare (http://www.isabelhealthcare.com). The current version of Isabel has two

components. They are the Isabel PRO Diagnosis Reminder System (IDRS) and the Isabel

PRO Knowledge Mobilizing System.

Isabel PRO Diagnosis Reminder System (IDRS)

The IDRS system works by a provider entering in the clinical symptoms of the

encounter and other patient demographic information into the system. The system then

uses natural language processing and search algorithms to search the Isabel database for

diagnoses that match the clinical information entered. The clinical diagnoses returned are

not ranked based on clinical probability. Isabel does this because the application is

designed to be more of a check list rather than an expert system that suggests diagnoses.

Isabel does however breakout the clinical diagnoses based on body systems. If a specific

diagnosis is clicked on then the application displays sign-symptoms, tests, and other

information for that diagnosis. The application can also be configured to show the

diagnostic list in a predefined way like for example the system can be set to show the

most common diagnoses. In addition, Isabel has a module where the system can show the

medications that might cause the symptoms that are entered. Isabel also has a module





where the system can shows the Bioterrorism issues that might have caused the

symptoms that are entered.

A group of researchers tested the Isabel application using test cases, and Isabel

suggested the correct diagnosis in 48 of 50 cases (96%) with key findings entry. Isabel

was able to suggest the correct diagnosis in 37 of the 50 cases (74%) if the entire case

history was pasted in.

In order to evaluate the Isabel system the project case information from the May 5

outpatient visit was entered into Isabel PRO 2010 demo application. The clinical

information below was entered into the system

Age: adult 50-65

Gender: male

Region: American North

Specialty: General

Query text: numbness and tingling in left hand, numbness and tingling in left leg, Mild

weakness in upper extremities, Leg numbness in thigh region, lower left leg weakness

The application returned 40 diagnoses which were ordered by degree of match between

the query entered and the Isabel database.

In the nervous body system disorders category 13 clinical diagnosis are listed. They are

listed below.

Nervous System Disorders

1. Diabetic Neuropathy

2. Transient Ischemic Attack

3. Myelopathy

4. Spinal Cord Tumors

5. Cervical Spondylosis

6. Syringomyelia

7. Lumbar Disc Protrusion

8. Spinal Cord Injuries

9. Brain Neoplasms

10. Myelitis

11. Vascular Dementia

12. Paroxysmal Dyskinesias

13. Multiple Sclerosis

See Isabel demo application picture on the next page.





Sample diagnostic tool:

There a number of issues in using the Isabel diagnostic system that has been

identified by providers. Isabel does not allow the user to further limit the number of

diagnosis returned by entering in negative data. For example in the project case the

provider would not have been able to enter that patient reported no neck or back pain.

Isabel Healthcare is working on adding the feature of entering in negative data to future

versions. Sometimes Isabel can return too many diagnoses which make the list not very

useable by providers. In order to help with this issue Isabel Healthcare recommends

entering in medical terms and not values like numbers. Isabel Health also recommends

entering in at least 3-5 clinical information values in order to get the best results out of

the application.

Isabel PRO Knowledge Mobilizing System IKMS

IKMS is system that allows providers to search a database for clinical reference

information. The clinical information can be images, journal abstracts, information in

textbooks, and other information.





Other Isabel features

Isabel is designed to be used with a PDA and Isabel can be linked to applications

like UpToDate and MD Consult. Isabel can be integrated into EHR systems. Some of the

systems that Isabel can be integrated with are NextGen, PatientKeeper A4 Health

Systems, and a few hospital EMR vendors like Cerner. Isabel can be integrated into the

EHR so that information entered into EHR modules can be used to create a query string

that can be passed to Isabel. For example with NextGen information entered into the

reason for visit and history of present illness areas can be used to create a query string

that is automatically passed to Isabel. Please see picture below.

Sample EMR and diagnostic tool integration:





DxPlain

Information from the May 5 outpatient visit was also entered into DxPlain which

is another diagnostic decision support system. The current findings from DxPlain are

listed below.

1. Extremity muscle weakness, lower

2. Extremity muscle weakness, upper

3. Lower extremity numbness

4. Hand numbness

5. Male

6. Middle age (41 to 70 yrs

The common diseases with sufficient evidence to support the DX generated by DXPlain

are listed below.

(1) + Sclerosis, multiple

(2) + Spinal cord, compression

(3) + Carpal tunnel syndrome

(4) + Herniated cervical disc

(5) + Spondylosis, cervical

Differential diagnosis

Both DXPlain and Isable listed spinal cord diagnoses, cervical spine diagnoses,

and Multiple Sclerosis. The Isabel also listed a clinical diagnosis of Brain Neoplasm. MS

and Brain Neoplasm are rare diseases in the general population and based on the

descriptions of those diseases in DxPlain the patient does not have some the symptoms

related to those diseases. For example the patient has not complained of headaches which

are common symptom of a Brain Neoplasm. More than one of the patient’s symptoms

point to a spinal cord or cervical spine problem, but the patient says he has no back or

neck pain. The patient did claim of shoulder and neck pain in a prior encounter. In order

to rule out some of these possible diagnoses and help identify the correct diagnosis we

need to use a test to clear up the diagnostic uncertainty. The test selected will be

discussed later in the document.

Problem List

The classes of items that are appropriate to include in a patient’s problem list are

clinical diagnoses, syndromes, pathophysiologic states, cluster of clues, isolated

abnormalities, and pyschosocioecomonic issues. In the project case the clinical diagnosis

after the May 5 outpatient visit has not yet been determined. The patient does however

have clues and pyschosocioecomonic information that can be entered into the problem

List. The items in the problem list are listed below.





1. Numbness and tingling in left hand

2. Numbness and tingling in left leg

3. Mild weakness in upper extremities

4. Leg numbness in thigh region

5. Lower left leg weakness.

6. Unemployed

7. Smoker since 18

8. Prior street Drug use

The history of assessment section shows problems from patient encounters and is

automatically populated from information entered in the EHR by providers during the

encounter. The Chronic condition section shows chronic conditions like diabetes and

hypertension. Both areas can be populated with ICD9 codes. See picture below






Plan

Labs / Diagnostic Ordering

MRI was the only suggested diagnostic procedure. There was not enough

information to calculate sensitivity and specificity ratios.

An MRI of the neck was ordered to rule out significant pathology and spinal cord

compression. The automation of MRI ordering and processing is shown in the diagram

below.

Hospital Information System/PACS Workflow

RadiologyInformationSystem

Order EntrySystem

Hospital InformationSystem

ADT

ADTADT

ORDERS

ORDERS

ORDER RESULTS

ModalityWorklist

DICOM Images

ImagesReports

Validation, Study Updates, Patient Updates, Pre-fetching, Reports

Modality

Workstation

Server / Deep Archive

PACSBroker





The diagram below shows at the high level the flow applicable to our clinical case.

Study Read Events

Server

PACS Broker

Radiology InformationSystem

Broker Can Send RadiologyInformation Systeman HL7 ORM Message toIndicate a “Read” Status

HL7

Workstation

Workstation Updates Status on Server

Workstation UpdatesStudy Status to“Read”

DICOM

Report is Markedas Read

The results are electronically shared and captured in an EMR tool thus closing the loop

on diagnostic ordering.






Automated Patient Follow-up Process

The patient did not keep his follow-up appointment and there was no recorded

attempt to reach him. An improvement of the process would be to have automate steps in

following up with patients. Possible improvements are listed below:

• Automate identification of high risk patients. Integrate into EMR or other

information system in use. Allow for flags to be set by the system or manually.

• Reminder to the patient via phone (if not in place already). Automate notifications

via system alerts.

• Follow up via phone with critical patients in case of a missed appointment.

Automate notifications via system alerts. When no response, then automatically

reschedule and notify the patient.

• Produce reports of missed appointments on regular bases (weekly or monthly).

Make follow up calls to the critical patients with missed appointments.

A clear advantage of automated follow-ups is that the extended care to the patients,

closed loop on the individual case, and opened doors for proactive actions.

Improvement in Communicating Diagnoses to the Patient

One improvement to the process used in the case is to make available the diagnoses

and interpretation of all results via a web portal.

Another improvement is to enable patients to be better informed by making them

aware of and providing them with links to various resources available online.

Conclusion

Group 1 applied the concepts and tools taught in class and documented

improvements in the flow of information. There are numerous advantages in automating

the process for collecting and processing the data, as well as those for communicating

with the patient and following up. Keeping all relevant patient information in an EMR

system, using tools to communicate the results of the physical exam, utilizing diagnostic

tools and other resources to work the diagnoses, automating ordering and processing of

MRI, alerting for follow up calls, and enabling easy access to results, interpretations and

additional resources to patients improves the flow and use of available information. In

this case however, the process improvements aided by the new technology would have

little impact on the outcome.





References

Graber Mark,Mathew Ashlei (2008).

Performance of a Web-Based Clinical Diagnosis Support System for Internists. Journal of

General Internal Medicine; Issue Volume 23, Supplement 1,37-40

Improving Patient Care and Safety: Use of electronic diagnosis reminder systems (2009),

http://www.isabelhealthcare.com/home/demo_new

Mack EH, Wheeler DS, Embi PJ. (2009),

Clinical decision support systems in the pediatric intensive care unit, Pediatr Crit Care Med Vol.

10, No. 1 23-28

http://www.isabelhealthcare.com/home/demo_new

Documents

Final Group Presentation, Summary Document162389445480246740.weebly.com/uploads/1/4/5/9/... · Presentation Date: March 1, 2010 Page: 1 of 15 Final Group Presentation, Summary Document