Closing the Evidence Gap
Doctors, pharmacists, naturopaths, nurses, and other health professionals are able to write prescriptions within a regulatory framework identified by their professional organizations. All these bodies have remarkably similar approaches to prescribing as one might expect. They are based on recommendations set out more than 60 years ago 1,2 when the knowledge about compounding of drugs was as important as the clinical evidence about the drug itself.
When prescribing, health professionals consider these three items:
- The patient’s wishes in terms of desired benefits for the condition, and acceptability of risk of harm.
- The clinical evidence for benefit and harm of the potential medication options for the condition.
- Adjustment of those drug options by taking into account the other conditions the patient may have, the current state of those conditions, and the biophysical status that may affect the absorption, metabolism, distribution, and excretion of the medications.
The patient-centered approach and shared decision making are part of what every health professional is taught and has been taught for over 50 years 3. Tools have been developed to aid this process with some success though they are still not used extensively, probably because of the second and third items being such a challenge 4.
Sixty years ago, assessing the clinical evidence was not a challenge when there were less than 100 drugs used in practice. Today there are over 1,400 drugs approved by the FDA alone. Physicians need more than 2,000,000 pieces of information, including clinical evidence, to help manage patients 5. This may seem hard to believe but for even one condition such as asthma, there were 67 randomized controlled trials for effectiveness published in 2016 alone. The amount of clinical evidence was doubling at the rate of every seven years in 1980, but by 2010 it was every three and a half years. “Knowledge is expanding faster than our ability to assimilate and apply it effectively” 6. It has been estimated that a primary care physician would need to read for 627.5 hours per month to cover the 341 journal articles with material relevant to their specialty 7.
To address limitations of memory, or time to read new evidence, many physicians will look up databases for new evidence 8 around or during the time of the consultation 9. Clinical uncertainty is identified by clinicians for every other patient they see. They only pursue half of these uncertainties finding answers to 78% of them. This leaves a considerable clinical evidence gap and one-third of the evidence gap is about drug treatment 10.
The integration of multiple patient variables, to identify or modify drug and dosage options, is very difficult for patients with several conditions taking multiple medications. The number of people with multiple conditions has increased with rates of 13% of Canadians with two or more chronic conditions. For example, impaired kidney function may impact dosing for drugs used in the treatment of hypertension, diabetes, arthritis, and many other conditions. Each drug is affected differently by varying degrees of renal impairment and so the prescriber needs to be aware of which drugs are affected and by how much the dose needs changing.
In summary, there is too much evidence for a health professional to have at their fingertips, integration of patient variables is highly complex, and there is too little time to do this effectively in the normal consultation.
Does this matter?
The lack of knowledge at the time of prescribing may lead to an inappropriate prescription that is either ineffective, harmful or both. The rate of potentially inappropriate prescribing is now greater than 40% 11. The result of an inappropriate prescription may be an unexpected and harmful adverse drug reaction.
Adverse drug reactions are very common with ten percent of patients in primary care reporting having an adverse drug reaction within the previous six months 12. Adverse drug reactions are the fourth cause of death in the USA (FDA) responsible for as many as 500,000 deaths worldwide every year. Two significant mechanisms have been introduced to try and address this.
Designed to address the clinical evidence gap of health professionals, guidelines for disease management were introduced over the last thirty years. There are two issues with guidelines. They generally summarise the evidence very effectively but only for one disease at a time. This means that they do not integrate their recommendations with other disease guidelines and so may conflict with other guideline recommendations. For example, in the guideline for diabetes, there were there were 133 potentially serious interactions with drugs recommended by other disease guidelines 13. The second issue is that they are designed for the average patient. One cannot integrate individual patient data into the guideline to see how this alters the guideline recommendations 14. Dissemination and uptake of guidelines into practice remain a challenge despite multiple approaches 15.
The second innovation was the introduction of computerized alerts based on drug-drug interaction databases. Many electronic medical records have drug alert systems that let the physician know if there is a potential interaction. Partly because these alerts only occur at the end of consultation most alerts are overridden by as much as 90% of the time 16.
Given the massive amount of clinical evidence and patient information required to identify the list of potential safe and effective medications, it is not surprising that we see such high rates of potentially inappropriate prescriptions and adverse drug reactions. Despite guidelines and drug alerts we still have what some describe as an epidemic of prescribing related deaths due to adverse drug reactions. Recording of basic data in the electronic records is one step, but there remains the challenge of integration of all this evidence at the time of identifying optimal drug options.
Addressing major challenges of prescribing with the TreatGx decision support tool
Our team recognized that to address the two major challenges of prescribing we needed to have a system that could integrate patient variables with the clinical evidence of medication effectiveness and harm. The clinical evidence can be identified and synthesized so that it can be used to make smart algorithms. We have done this for over 30 conditions in the TreatGx decision support system.
The decision support system combines patient information with the clinical evidence to identify the optimal drug options, dose adjusted, for that individual. This list of options can be shared with the patient. The prescriber selects the disease and enters relevant patient variables. Many of the variables can be taken straight from the electronic medical record and the system will even incorporate genetic information if available. The software identifies the drug options, dose adjusted, with information relevant to that patient.
By combining patient information with clinical evidence TreatGx software is enabling an evidence-informed approach to personalized care.
Martin Dawes
Founder and CEO of GenXys Health Care Systems
[button size=’small’ style=” text= ‘More information for health care providers’ icon=” icon_color=” link=’/content/how-it-works-hcp/’ target=’_self’ color=” hover_color=” border_color=” hover_border_color=” background_color=” hover_background_color=” font_style=” font_weight=” text_align=” margin=”] [button size=’small’ style=” text= ‘Medications covered’ icon=” icon_color=” link=’/content/conditions-and-medications/’ target=’_self’ color=” hover_color=” border_color=” hover_border_color=” background_color=” hover_background_color=” font_style=” font_weight=” text_align=” margin=”]
References
1. The Medical Act, 1950. Br Med J. BMJ Publishing Group; 1950 Aug 5;2(4674):337–8.
2. DARLINGTON RC. A procedure for improving prescription writing. J Natl Med Assoc. National Medical Association; 1955 Sep;47(5):297–307.
3. Balint E. The possibilities of patient-centered medicine. J R Coll Gen Pract. Royal College of General Practitioners; 1969 May;17(82):269–76.
4. Fried TR, Tinetti ME, Iannone L, O’Leary JR, Towle V, Van Ness PH. Health Outcome Prioritization as a Tool for Decision Making Among Older Persons With Multiple Chronic Conditions. Arch Intern Med. American Medical Association; 2011;171(20):1854–6.
5. Pauker SG, Gorry GA, Kassirer JP, Schwartz WB. Towards the simulation of clinical cognition. Taking a present illness by computer. The American journal of medicine. 1976 Jun;60(7):981–96.
6. Densen P. Challenges and opportunities facing medical education. Trans Am Clin Climatol Assoc. American Clinical and Climatological Association; 2011;122:48–58.
7. Alper BS, Hand JA, Elliott SG, Kinkade S, Hauan MJ, Onion DK, et al. How much effort is needed to keep up with the literature relevant for primary care? J Med Libr Assoc. 2004 Oct;92(4):429–37.
8. González-González AI, Dawes M, Sánchez-Mateos J, Riesgo-Fuertes R, Escortell-Mayor E, Sanz-Cuesta T, et al. Information needs and information-seeking behavior of primary care physicians. Ann Fam Med. 2007 Jun;5(4):345–52.
9. Ely J, Osheroff J, Ebell M, Bergus G, Levy B, Chambliss M, et al. Analysis of questions asked by family doctors regarding patient care. BMJ. 1999;319:358–61.
10. Del Fiol G, Workman TE, Gorman PN. Clinical questions raised by clinicians at the point of care: a systematic review. JAMA Intern Med. American Medical Association; 2014 May;174(5):710–8.
11. Moriarty F, Cahir C, Bennett K, Hughes CM, Kenny RA, Fahey T. Potentially inappropriate prescribing and its association with health outcomes in middle-aged people: a prospective cohort study in Ireland. BMJ Open. British Medical Journal Publishing Group; 2017 Oct 16;7(10):e016562.
12. Miller GC, Britt HC, Valenti L. Adverse drug events in general practice patients in Australia. Medical Journal of Australia. 2006.
13. Dumbreck S, Flynn A, Nairn M, Wilson M, Treweek S, Mercer SW, et al. Drug-disease and drug-drug interactions: systematic examination of recommendations in 12 UK national clinical guidelines. BMJ. BMJ Publishing Group Ltd; 2015;350(mar11 2):h949–9.
14. Dawes M. Co-morbidity: we need a guideline for each patient not a guideline for each disease. Fam Pract. 2010 Feb;27(1):1–2.
15. Shanbhag D, Graham ID, Harlos K, Haynes RB, Gabizon I, Connolly SJ, et al. Effectiveness of implementation interventions in improving physician adherence to guideline recommendations in heart failure: a systematic review. BMJ Open. British Medical Journal Publishing Group; 2018 Mar 6;8(3):e017765.
16. Isaac T, Weissman JS, Davis RB, Massagli M, Cyrulik A, Sands DZ, et al. Overrides of Medication Alerts in Ambulatory Care. Arch Intern Med. American Medical Association; 2009 Feb 9;169(3):305–11.