Incorporating patient experience and preferences into the paradigm of EBM represents a significant challenge. Similarly, critically evaluating publications that take into account this element is crucial but also presents its own difficulties.
In this context, it is essential to recognize that EBM seeks to integrate the best available scientific evidence with clinical experience and patient values and preferences to make informed and personalized decisions in healthcare. However, achieving this integration effectively involves overcoming various obstacles.
On one hand, including patient experience and preferences requires a more holistic approach to data collection and analysis, which is often not adequately addressed in traditional research. Additionally, variability in individual preferences and values can hinder the generalization of study results to clinical practice.
On the other hand, critically evaluating publications in this context is challenged by the need to identify and assess the quality of evidence related to patient experience and preferences. This involves considering the validity and relevance of studies, as well as how qualitative and quantitative aspects of this information have been integrated into the interpretation of results.
There is a "CliffsNotes" approach to understanding a medical study/journal report. It is useful to read the Methods section in order to see how the patients were selected and how the end points are defined. Then go right to the Results section and look at the raw data and the percentage differences in the incidences for each end point. Ignore the statistical analysis with "p" values, confidence intervals, etc. Then use your common sense to decide whether the differences are of any practical significance. Always keep in mind the law of small numbers: small sample size and/or low incidence of end points renders differences in incidence far less reliable. Most people have an intuitive awareness of the law of large numbers but often fail to understand the opposite. This shortcut can give one an immediate sense of whether the study is worth the time and effort to analyze it in more detail.
I second your points: was patient selection “clean” or were there uncontrolled variables that could sabotage the outcome? Are the end points clearly/simply defined or could there be too much “noise?” Common sense and your clinical experience applied to the raw data will allow you to come to a quick, but usually sensible decision on whether or not to apply the results to your patients.
The questions you added certainly need to go into any critical appraisal of a scientific study. Another (and, possibly the most important) is the quality of the data being analyzed. For example, mortality figures are probably reasonably accurate, but cause of death figures are not. These usually come from death certificates that are notoriously unreliable sources of information on cause of death. Accepting anything on a death certificate besides the fact of death is unwarranted. The recent covid fiasco is another example. Not only was there never a scientifically acceptable isolation of a specific virus, but the so called "tests" for it were wildly inaccurate. So all the figures given about incidence, transmission, asymptomatic cases, "long" covid, etc. are all pure nonsense and can be disregarded even before considering statistical analysis.
My general policy is to always ask "Are you sure?" to everything that ever gets put forward to me. Regardless of whether I'm likely to agree with the ideas being put forward. I was recently having a discussion about the idea of artificial intelligence boils down to one simple principle.
Perhaps some wisdom on how to make this happen in frontline medicine. I am grateful for the review of this important concept, but find that most of my colleagues would like a fairly straightforward approach to treating patients as they present in the clinic or ED. Not many are afforded the time to “stop and think “ ( hat tip John Mandrola) in day to day practice. Perhaps AI and massive computing power will save the day!
Wow this brings back memories. We went through that JAMA series as medical residents in the late 90s.
Also good to have that explicit reminder: as refined as you can get with the evidence, that’s just 1 of the 3 aspects one needs to consider in the course of clinical decision making.
I shared this post with one of my brilliant co-residents during sign-out this morning. Always feeling grateful for your work! Cannot wait to see what's next.
I'm looking forward to finding out whether you are an authentic Bayesian, a semi-Bayesian, or a faux Bayesian. I'm also especially looking forward to learning what you think of the various laughable tricks used to "spruce up" purely observational data interpretation so that promotion-seeking assistant professors can "grow the C.V." by publishing lots of badly-written papers that plant seeds of belief in various sketchy cause-effect mechanisms and always end up spewing the tired cliche, "more work needs to be done".
The only time that critical skills will be required is if there is a preponderance of factual and empirical data and studies from which to draw realistic conclusions. We are extremely far from that in the field of medicine.
I just retired from active practice last week after 25 years, so of course this would help refresh my knowledge. Let's face it, you tend to get into a rut of how you practice over time.
Incorporating patient experience and preferences into the paradigm of EBM represents a significant challenge. Similarly, critically evaluating publications that take into account this element is crucial but also presents its own difficulties.
In this context, it is essential to recognize that EBM seeks to integrate the best available scientific evidence with clinical experience and patient values and preferences to make informed and personalized decisions in healthcare. However, achieving this integration effectively involves overcoming various obstacles.
On one hand, including patient experience and preferences requires a more holistic approach to data collection and analysis, which is often not adequately addressed in traditional research. Additionally, variability in individual preferences and values can hinder the generalization of study results to clinical practice.
On the other hand, critically evaluating publications in this context is challenged by the need to identify and assess the quality of evidence related to patient experience and preferences. This involves considering the validity and relevance of studies, as well as how qualitative and quantitative aspects of this information have been integrated into the interpretation of results.
Translating to English:
I am excited to start reading to them.
Dayami Martinez
There is a "CliffsNotes" approach to understanding a medical study/journal report. It is useful to read the Methods section in order to see how the patients were selected and how the end points are defined. Then go right to the Results section and look at the raw data and the percentage differences in the incidences for each end point. Ignore the statistical analysis with "p" values, confidence intervals, etc. Then use your common sense to decide whether the differences are of any practical significance. Always keep in mind the law of small numbers: small sample size and/or low incidence of end points renders differences in incidence far less reliable. Most people have an intuitive awareness of the law of large numbers but often fail to understand the opposite. This shortcut can give one an immediate sense of whether the study is worth the time and effort to analyze it in more detail.
I second your points: was patient selection “clean” or were there uncontrolled variables that could sabotage the outcome? Are the end points clearly/simply defined or could there be too much “noise?” Common sense and your clinical experience applied to the raw data will allow you to come to a quick, but usually sensible decision on whether or not to apply the results to your patients.
The questions you added certainly need to go into any critical appraisal of a scientific study. Another (and, possibly the most important) is the quality of the data being analyzed. For example, mortality figures are probably reasonably accurate, but cause of death figures are not. These usually come from death certificates that are notoriously unreliable sources of information on cause of death. Accepting anything on a death certificate besides the fact of death is unwarranted. The recent covid fiasco is another example. Not only was there never a scientifically acceptable isolation of a specific virus, but the so called "tests" for it were wildly inaccurate. So all the figures given about incidence, transmission, asymptomatic cases, "long" covid, etc. are all pure nonsense and can be disregarded even before considering statistical analysis.
My general policy is to always ask "Are you sure?" to everything that ever gets put forward to me. Regardless of whether I'm likely to agree with the ideas being put forward. I was recently having a discussion about the idea of artificial intelligence boils down to one simple principle.
"Be humble about the things you don't know."
Perhaps some wisdom on how to make this happen in frontline medicine. I am grateful for the review of this important concept, but find that most of my colleagues would like a fairly straightforward approach to treating patients as they present in the clinic or ED. Not many are afforded the time to “stop and think “ ( hat tip John Mandrola) in day to day practice. Perhaps AI and massive computing power will save the day!
Wow this brings back memories. We went through that JAMA series as medical residents in the late 90s.
Also good to have that explicit reminder: as refined as you can get with the evidence, that’s just 1 of the 3 aspects one needs to consider in the course of clinical decision making.
Oh this will be fun.
I shared this post with one of my brilliant co-residents during sign-out this morning. Always feeling grateful for your work! Cannot wait to see what's next.
I'm looking forward to finding out whether you are an authentic Bayesian, a semi-Bayesian, or a faux Bayesian. I'm also especially looking forward to learning what you think of the various laughable tricks used to "spruce up" purely observational data interpretation so that promotion-seeking assistant professors can "grow the C.V." by publishing lots of badly-written papers that plant seeds of belief in various sketchy cause-effect mechanisms and always end up spewing the tired cliche, "more work needs to be done".
Now I know to work to hide my Bayesian credentials.
The only time that critical skills will be required is if there is a preponderance of factual and empirical data and studies from which to draw realistic conclusions. We are extremely far from that in the field of medicine.
I just retired from active practice last week after 25 years, so of course this would help refresh my knowledge. Let's face it, you tend to get into a rut of how you practice over time.
Fantastic. Re-upped my lapsed paid subscription for this. Looking forward!
Well, it was about the time! “Give a Man a Fish, and You Feed Him for a Day. Teach a Man To Fish, and You Feed Him for a Lifetime”.