THRESHOLD

On a listserv discussing evidence-based healthcare there were comments arguing for and against using THRESHOLDs for decision-making. It prompted me to share the following:

An evidence-based, SHARED DECISION MAKING, patient-centered, ethically sound approach to treatment decisions (yes vs. no, or comparing options) can be made WITHOUT THRESHOLDS.

For example, for a decision to use or not use a treatment to reduce the risk of fragility fractures, let’s assume a valid summary of the relevant outcomes of importance to your patient is:

·        She has a 10% risk of fragility fracture in the next 10 years without the treatment and the treatment will reduce her risk to 7%, but

·        8% of the people who take the treatment will have heartburn.

This data can be shared with the patient in various forms and an informed decision can incorporate the patient’s relative valuation of the trade-off between benefits and harms.

The patient does not need an explicit THRESHOLD to make this decision, and the healthcare professional does not need an explicit THRESHOLD to inform this decision. Although one can use mathematical approaches to weigh the benefit vs. harm and report if it results in a NET BENEFIT (i.e. crosses a THRESHOLD on the favorable side) one does not need to.  We make decisions all the time without explicit THRESHOLDs.  Is buying a bottle of water worth the price?  Did you ever calculate the per-milliliter price to determine if it crosses a THRESHOLD to answer that question for yourself?

There are many USES OF THRESHOLDS in healthcare decision making, all of which are essentially transferring some aspect of decision making to a third party.

When considering cost, cost-effectiveness, cost-benefit, then a THRESHOLD may be applied.  This usually occurs when the purchaser is not the consumer. Whether a government, insurance company, or other group is paying for the healthcare services they may apply financial stewardship decisions and suggest or impose a THRESHOLD for this reason.  This may or may not be included in clinical practice guidelines.

When considering net benefit or net harm to people other than the patient, there is no way to ask the other people at the moment of decision making. So, we may apply THRESHOLD for decisions that can harm others, such as reducing risk of infection or injurious behaviors.

When the patient is not competent to participate in informed decision making we may need THRESHOLD.

When we define a diagnosis using THRESHOLD, and that diagnosis changes the eligibility for the decision making, we may be using the diagnostic classification as one of the reasons for THRESHOLD noted above OR we may be using it to match our “information” to inform decision making with the available evidence.  This latter situation (where THRESHOLDS are used for diagnostic classification) is where false positives and false negatives occur and that adds another set of harms or risks to consider.

Have said all that, we do not need THRESHOLDS for most of our decision-making but we still look to use THRESHOLDS because…

As individuals facing the complexity and responsibility of informing healthcare decision making we may seek THRESHOLD because they are EASIER than making individualized decisions.  This may not be better for the decisions but may be needed to cope with the high volume of care and high complexity until we have tools that make it EASY to see just what we need, just when we need it for such decision making support.