Why Single Disease Guidelines Break Down in People With Many Conditions

Single disease guidelines break down in multimorbidity because each one is written as if its condition were the only one a person has. Stack a diabetes guideline on top of a heart failure guideline on top of an osteoporosis guideline, and you get a schedule of pills, tests, and appointments that no single author ever designed or intended. The National Institute for Health and Care Excellence addressed this in guideline NG56, "Multimorbidity: clinical assessment and management," published in September 2016. Its central move is to treat the total burden of care as a clinical problem in its own right, and to ask which treatments actually match what a given person wants from their remaining years.

How single-condition guidelines are built

Most clinical guidelines start with the evidence from randomized trials, and most trials recruit relatively uncomplicated participants. Someone with heart failure plus kidney disease plus depression plus early dementia is often screened out, because those other conditions make the data harder to interpret. The evidence in single-condition guidance is therefore drawn largely from people who do not carry the multimorbidity, frailty, or medicine counts of the patient sitting in front of a clinician. A guideline can be internally rigorous and still rest on a population that looks nothing like the people it ends up being applied to.

That gap matters because multimorbidity is the common case, not the exception. In their British Journal of General Practice review of NG56, David Kernick, Carolyn Chew-Graham, and Norma O'Flynn note that roughly two-thirds of people over 65 have two or more long-term conditions. Guidelines optimized for the person with one disease are, for older adults, optimized for a minority.

The arithmetic of stacking

The failure is additive. Each guideline, taken alone, is defensible. The problem appears when they are summed. A person may reasonably be advised to take a statin, an antihypertensive or two, a diabetes agent, an inhaler, a bone-protection drug, an antidepressant, and a proton pump inhibitor to cover the side effects of the others, each with its own monitoring bloods and follow-up visit. No one prescribed all of that at once. It accreted, one sensible recommendation at a time.

NG56 puts a number on where this becomes a flag for review. It advises clinicians to consider a multimorbidity approach for anyone prescribed 10 to 14 regular medicines, and to do so automatically for anyone on 15 or more. Those thresholds are not claims that many medicines are inherently wrong. They are a signal that the sum of individually correct decisions has grown large enough to deserve a look as a whole.

What treatment burden actually means

The concept doing the work in NG56 is treatment burden: the total demand that managing illness places on a person's time, attention, and daily life. It includes the medicines themselves, the appointments, the monitoring, the dietary rules, the coordination between services that do not talk to each other, and the mental load of keeping all of it straight. A regimen can be pharmacologically optimal and still be unlivable.

The guideline is candid that a large share of this burden comes from managing risk factors for future disease rather than treating present symptoms. Much of preventive medicine asks people to accept a real cost today (a daily pill, a side effect, a clinic visit) for a statistical reduction in a bad event years from now. For a person with many conditions and a shorter horizon, that trade can shift. NG56 asks clinicians to identify treatments with high burden, limited benefit, or meaningful risk of harm, and to raise them for honest discussion.

Reasoning about competing priorities

The alternative NG56 proposes is not a rival algorithm that spits out a different pill list. It is a method for weighing priorities that only the patient can rank. The guideline asks clinicians to establish what matters to each person, whether that is independence, staying in work, avoiding a specific outcome such as stroke, reducing the harms of medication, or extending life, and then to tailor recommendations to that goal rather than to any one disease.

From there, review becomes structured. NG56 supports discussing whether to stop treatments with no clear benefit, whether symptom drugs are causing more harm than relief, and whether preventive treatments still make sense when life expectancy is limited or frailty is advanced. It uses bone-protection bisphosphonates as a worked example, suggesting a conversation about stopping after about three years in some patients. The logic generalizes: a treatment whose benefit lands a decade out is a different proposition depending on what the next decade looks like.

Who a multimorbidity approach is for

NG56 does not ask clinicians to run every patient through this process. It names the people most likely to benefit: those who find their treatments or daily activities hard to manage, those receiving fragmented care from many services, those with both physical and mental health conditions, those living with frailty, and those with frequent unplanned hospital admissions. It points to validated tools, such as the electronic frailty index and admission-risk scores like QAdmissions, to help identify who is at risk, alongside simply asking.

For prevention specifically, the lesson is not that guidelines are wrong or that risk-factor treatment does not work. It is that a recommendation validated in people with one condition carries an unstated assumption of simplicity that vanishes in real patients. Reading a guideline well means reading who it was built from. This article is educational and is not medical advice; decisions about starting or stopping any treatment belong with a person and their own clinician.