A medical algorithm (medical calculator) that is used by itself can solve a specific problem. However, medical algorithms must be used in concert with many others in order to achieve their true potential. We call such a collection of medical algorithms a “MedalPack.” Each MedalPack is intended to address a specific clinical problem.
Two Types of Clinical Problems
Clinical problems can be categorized as either specific or nonspecific.
Nonspecific Clinical Problem
A nonspecific problem is a symptom or situation that may be caused by several disorders. These problems are ambiguous and the goal is to refine the problem until a specific cause can be identified.
An example of a MedalPack dealing with a nonspecific clinical problem is Patient Safety. Here the issue is protecting the patient from many kinds of hazards. Some hazards are complications of therapy (surgery, drugs, cancer care) while others are associated with the healthcare environment in general (infection control, physical and environmental hazards). A few samples of each algorithm type include:
Complications of Therapy:
- Medication Regimen Complexity Index (MRCI)
- Prescription Abbreviations Associated with Dosage Errors
- Safety Guidelines for Patients Taking Chemotherapy Agents at Home
- Warfarin Resistance Associated with Poor Intestinal Absorption
- Preventing Errors in Medical Gas Administration
Healthcare Environment Hazards:
- Barium Embolism Following Barium Enema
- Clinical Features of Allergic Flare Reaction to Intravenous Infusion of Drug
- Local Tissue Injury Associated with Extravasation of Contrast Material
- Heat Loss After Infusion of Refrigerated Blood
- Healthcare Provider Substance Abuse Warning Signs
Specific Clinical Problem
A specific problem deals with one condition and can be broken down into four basic tasks:
- Addressing patient-specific concerns.
An example of a MedalPack dealing with a specific disorder is Type II Diabetes. There are a number of risk factors for diabetes and the diagnosis can be made by specific criteria that may need to be modified based on age, race and gender. Once diabetes is diagnosed the patient faces a number of complications, which taken together with comorbid conditions determine the prognosis. Based on prognosis a therapy must be planned. Monitoring the patient over time will find whether the strategy is working or needs to improved and whether the patient is developing adverse effects. The success of any management depends on how the patient and disease interact over time.
When a person has one specific problem then diagnosis management is relatively straightforward. As the number of problems increases the complexity increases exponentially. For a patient with several chronic conditions it is not uncommon for him or her to be receiving over 15 different medications. When many drugs are being taken by a patient with multiple organ systems affected by disease then prediction of outcomes may no longer follow simple heuristics.
When a busy clinician encounters a complex patient the issues can be numerous and overwhelming. At the same time the allotted time to be with a patient decreases and the demands for documentation increase. Either the patient has to be seen by multiple specialists with loss of coherent oversight or the doctor has to skimp on less critical problems. In the past it was the patient who suffered in the long run. With the changes in reimbursement that are currently underway, the provider may now be affected if quality goals are not met.
Sample algorithms that can assist physicians handling a type 2 diabetes patient with multiple complications are:
- Rebound Hyperglycemia Following Excessive Insulin (Somogyi Effect)
- Criteria for Small Fiber Neuropathy (SFN)
- Prognostic Factors of Community-Acquired Pneumonia in Patient with Diabetes Mellitus
- Peripheral Arterial Disease Score (PADS)
- Diabetic Macular Edema Disease Severity Scale
Role of Automated Medical Algorithms
Automated medical algorithms organized into MedalPacks offer a new option. These algorithms can run in the background registering hundreds of issues affecting the patient (hobbies, travels, nutrition, exercise, work, social interactions, etc) then summarizing and prioritizing them for the clinician. Medicine is both a science and art. Automated algorithms can consolidate the science, leaving an experienced clinician more time to practice her or his art for the patient.