Pain Assessment Instruments

Pain assessment instruments fall into four groups that answer different questions. Intensity scales such as the Numeric Pain Rating Scale and the Visual Analogue Scale ask how bad the pain is right now. Screening tools such as painDETECT, DN4 and LANSS ask whether the pain has a neuropathic component, which changes treatment. Interference measures such as the Brief Pain Inventory and the West Haven-Yale Multidimensional Pain Inventory ask what the pain stops the person doing. Observational tools such as PAINAD, the Abbey Pain Scale, the Behavioral Pain Scale and CRIES exist because the person cannot self-report at all.

Choosing between them starts with who is being assessed and what decision the score will inform. A single intensity number is enough to track response to an analgesic over an hour. It is not enough to judge whether a chronic pain programme worked, because intensity and interference move independently: people often report the same pain intensity while doing considerably more.

In clinical care, intensity scales are usually recorded at every contact and the longer instruments at intake and discharge. In research, the choice is tighter, because the primary endpoint has to be defined in advance and the instrument has to be sensitive to the change the study expects to produce.

Observational scales carry a specific caution. They score behaviour, not experience, and behaviour is affected by sedation, delirium and motor impairment as well as by pain. A rising PAINAD score in advanced dementia is a signal to look for a cause, not a measurement of how much it hurts. The same applies in intensive care, where the Behavioral Pain Scale and the Critical-Care Pain Observation Tool were developed precisely because ventilated patients cannot speak, and where sedation depth confounds every observed item.

Three things are worth checking before a pain instrument goes into a protocol. First, whether a minimal clinically important difference has been published for the population being studied. For the Numeric Pain Rating Scale a change of about two points is commonly cited, but the figure varies by condition and by baseline severity, and a study powered on the wrong number is a study that will not detect its own effect.

Second, whether the instrument was validated in the setting where it will be used. Screening tools for neuropathic pain were mostly developed in specialist clinics, where the pre-test probability is high. Their performance in primary care is not the same.

Third, licensing. Several widely used pain instruments are copyrighted and require permission or a fee for commercial or research use, and some translations are controlled separately from the original. A translated version is a different instrument until it has been revalidated in the target language, not simply the same questions in another wording.

Which pain scale should I use for a clinical trial?

It depends on the endpoint. If the endpoint is pain intensity, a single well-anchored numeric or visual analogue scale is usually the most sensitive and the easiest to collect repeatedly. If the endpoint is function or quality of life, an interference measure such as the Brief Pain Inventory answers the question better, and the two should not be swapped for each other during a study.

How do I assess pain in someone who cannot self-report?

Use an observational instrument matched to the population: PAINAD or the Abbey Pain Scale in advanced dementia, the Behavioral Pain Scale or the Critical-Care Pain Observation Tool in intensive care, CRIES or the Neonatal Infant Pain Scale in newborns, and FLACC in young children. Record what else could be driving the behaviour, because none of these tools can separate pain from agitation on their own.

Are pain scales free to use?

Some are and some are not. Simple intensity scales are generally free. Several multidimensional instruments are copyrighted, and permission depends on whether the use is clinical, academic or commercial. Check the licence with the copyright holder before a study starts rather than after data collection.

What is a meaningful change in a pain score?

Statistically significant and clinically meaningful are different things. Published minimal clinically important differences exist for many pain instruments, but they are population specific and depend on baseline severity. Use the value published for the closest matching population, and state which one was used.