Chapter 39 — Further Reading

This chapter's subject is unusual for this book: the primary literature is not about molecules at all. It is about communication, disclosure, shared decision-making, and — for the peri-operative material — a genuinely live area of professional guidance that has been revised since it first appeared and will be revised again.

A word on that last point before the list. The peri-operative guidance on GLP-1 receptor agonists is moving. The early recommendations were more categorical than the current ones, and any specific rule you read here will be out of date at some point. That is not a flaw in the guidance; it is what a young question looks like while it is being worked out. Read these sources for the reasoning, not for the rule, and remember §39.2's point: the reader's job is disclosure, and the rule is the anesthetist's job.

Sources are grouped by what they are for, not by difficulty alone.

Tier 1 — Start here

Professional society guidance on peri-operative management of GLP-1 receptor agonists. The American Society of Anesthesiologists issued the first widely circulated guidance in 2023; subsequent multi-society clinical practice guidance — developed jointly with gastroenterology, bariatric surgery, and related bodies — moved toward a more individualized assessment. Read the most recent version you can find rather than the first, and read the rationale sections rather than the recommendation boxes. What you are looking for is why the recommendation changed, which is a compact lesson in how a profession handles a new signal.

Any current review of gastric emptying effects across the GLP-1 receptor agonist class. The useful ones distinguish short-acting from long-acting agents and address whether the effect attenuates with continued exposure. This is the mechanistic background to §39.2 and to the 🧬 callout, and it explains why blanket rules were always going to be a poor fit.

The general literature on patient non-disclosure of complementary, alternative, and non-prescribed substance use. This is a large, old, and consistent body of work spanning many clinical settings. Read it for the reasons patients give — expectation of disapproval, belief that it was not relevant, and never having been asked in a way that invited an answer — which map directly onto §39.2's argument about why moral framing suppresses disclosure. Note that specific figures vary enormously by population and by how the question was asked; the pattern is robust, the numbers are not transferable.

A plain introduction to shared decision-making. Several national health bodies and professional colleges publish short, readable versions. The five questions of §39.4 are essentially the patient-side toolkit of this framework, and seeing the clinician-side version makes the encounter legible from both chairs.

Tier 2 — Go deeper

"Ask 3 Questions" / "AskShareKnow" style programs. Several health systems have developed and evaluated short question sets for patients — typically covering options, benefits and harms, and how likely each is. The evaluation literature on these is worth reading because it addresses the practical question this chapter cannot answer from first principles: does giving patients a question set actually change what happens in the room? The findings are modest and real, which is itself instructive.

Literature on the therapeutic relationship and continuity of care. §39.10's claim that a clinical relationship compounds is not merely rhetorical. There is a substantial body of work on continuity of care and its association with outcomes, and it is worth understanding both what that literature shows and where the causal inference is hard — patients who maintain continuity differ from those who do not in ways that are difficult to adjust for.

Anything serious on physician dispensing, self-referral, and financial conflicts of interest in clinical recommendation. This is the evidentiary basis for §39.8 and for Chapter 42's rating. The literature is more nuanced than either the credulous or the cynical reading suggests: conflicts are associated with measurable shifts in prescribing patterns, and this is compatible with almost every individual clinician believing exactly what they say. That combination is the thing to understand.

The pharmacist's scope of practice in your jurisdiction. Genuinely worth twenty minutes. §39.5 argues that pharmacists are the most underused clinical resource available to most patients, and what they are permitted to do — and what they are trained to do — varies more than people expect.

Chapter 28's nesiritide discussion, revisited. Re-read it specifically as an example of §39.6's second sentence. A compound with an excellent mechanistic story, an approved indication, and widespread enthusiasm, which an adequately powered trial did not support. "It doesn't work" is a strictly stronger claim than "we don't know," and this is what it looks like when it is earned.

Tier 3 — For the committed

Primary literature on communicating uncertainty in clinical encounters. There is a real research question underneath §39.6: how patients interpret expressions of uncertainty, and whether clinicians who express it are trusted more or less. The findings are not uniformly comforting, and they are directly relevant to the chapter's claim that "I don't know" is professionally expensive.

Work on the epistemics of clinical judgment. What kind of thing is a clinical judgment, how does it relate to evidence, and why is a prescription not a demonstration? Chapter 38 §38.10 and §39.8 both rest on an answer to this that the philosophy-of-medicine literature works through carefully.

Evidence on how medication reconciliation fails. §39.2's four failures are abstract until you read the operational literature on medication reconciliation at transitions of care — admission, discharge, transfer. That work quantifies how often the list is wrong and what happens when it is, and it makes the abstract argument concrete in a way nothing else does.

Case reports and case series on peri-operative aspiration in patients taking GLP-1 receptor agonists. Read these with Chapter 6's apparatus running: case reports are hypothesis-generating, subject to reporting bias in both directions, and cannot establish a rate. They are also how this signal was first noticed, which is a fair illustration of what case reports are genuinely good for.

The literature on placebo, expectation, and the therapeutic encounter itself. Case Study 39.2 notes that a fifty-minute consultation is a different experience from a fourteen-minute one, and that the difference is real. There is serious research on how much of a treatment effect is attributable to the encounter rather than the intervention. It is directly relevant to §39.8 and it is unsettling in both directions.

If you only do one thing

Write the five questions from §39.4 on your phone, and take them to your next appointment about anything.

Not a peptide appointment. Any appointment. A statin, a scan, a referral, a physical therapy recommendation, a change to something you have taken for years.

The point of doing it on something ordinary is that you will find out what the questions are like to ask when nothing is at stake and you are not invested — which is exactly the condition under which question 5 can still be asked honestly. You will also discover, usually, that your clinician answers them readily and that the appointment is better for it.

Then do the same thing with the one exposure you have not disclosed.