Selecting PROMs

PROMs can be used for many purposes. They are used in clinical practice, clinical trials, quality assurance activities, clinical registries, and benchmarking exercises. The same PROMs responses can be used for multiple purposes, especially when electronically accessible. At the beginning of a PROMs initiative, implementers must identify the purpose of collecting patients’ responses.

Generic and specific PROMs each have unique advantages, which need to be taken into consideration when choosing the most appropriate measures. To capitalise on the strengths of each type of PROM, a standard approach is to use generic measures with supplementary specific measures.

Generic PROMs measure aspects of health that most patients have in common. The advantage of generic PROMs is that they can be used across health conditions, treatments, and populations. Using generic measures, the relative impact of different health conditions and different treatments can be assessed.

Specific PROMs, on the other hand, are designed for particular health conditions or populations. The advantage of specific PROMs is that they tend to be more attuned to the concerns of the patients for whom they have been designed and typically are more sensitive to small, but clinically meaningful, changes in health.

Many PROMs have undergone rigorous psychometric development. Careful design and testing of PROMs increases confidence that they actually measure what they have been developed to measure (they are valid) and do so in a consistent manner (they are reliable). One indication of validity that can be particularly important is their responsiveness to change. Using PROMs that are more sensitive to changes over time can lead to a better understanding of the impacts of health conditions and associated treatments on patient outcomes.

There can be a temptation to modify PROMs through adding, deleting, or altering questions. Making such changes can invalidate the outputs from PROMs, reducing certainty that they still measure what they were designed to measure. Output from modified PROMs may not be able to be compared with output from unmodified PROMs, nor can it be reliably interpreted with reference to norms generated using the unmodified measures. When the wording of some questions are unsuitable for the target patient population (for example, some terms used in other countries may not be easily understood in Australia) the safest option may be to choose another PROM that does not have these issues. Alternatively, consulting a skilled psychometrician may be helpful in exploring other options, such as modifying and re-validating the PROM.

Measures of outcomes that are important to patients

Many PROMs have been developed in consultation with patients. When selecting PROMs for use, it is helpful to ensure that the measures reflect outcomes that are important to patients. Choosing PROMs with questions that resonate with the experiences and priorities of patients may encourage them to engage fully with the process of completing them and improve response rates. 

Appropriate for patient population

PROMs need to be appropriate for patient populations in terms of simplicity, accessibility, and length.¹ The instructions should be easily understood, the rating scales need to be intuitive, and the questions should be jargon-free. PROMs need to be accessible to everyone, including patients who communicate in languages other than English, patients with cognitive challenges, those with vision impairment, and others who may experience difficulties completing PROMs in the conventional ways. If PROMs are too long, there is a risk that they will not be completed. Clinical experience suggests that the time required to complete PROMs should not be more than 12 to 15 minutes.

Avoid duplication

Using generic and specific PROMs together is a powerful strategy, but introduces the possibility that some of the questions may be very similar. For example, some of the questions in a condition-specific health-related quality of life PROM may be duplicated in a generic health-related quality of life PROM. Patients can become irritated if they feel they are answering the same question several times.

Methods of collection

Patient responses to PROMs can be collected in various ways. Responses can be self-reported online, via tablet, or using pen and paper. PROMs can also be administered as structured interviews, either in person or over the telephone. Electronic methods have the advantage of making the capture, processing, and storage of patients’ responses more efficient.

Integration with clinical practice

Whichever method of administration is preferred, attention needs to be paid to how response collection and processing will be integrated into the regular workflows of clinical environments. Timely collection and processing of responses to allow PROMs to be used in clinical consultations can be a challenge, however there are increasingly novel ways of integrating PROMs into care. Some examples are provided in the case studies of PROMs implementation.

Resources for collection and processing 

Thought needs to be given to who will administer the PROMs. In some circumstances (eg. clinical quality registries), specialised support may be available to supervise administration and to manage processing. In circumstances where specialised assistance is not available, consideration should be given to involving other members of the clinical team. Electronic methods of administration – such as having patients complete PROMs on digital devices while waiting for their clinician or on their smartphone prior to the consultation – can facilitate collection and in some cases processing of PROMs. 

In clinical consultations

The outputs from PROMs should be presented in a way that patients and clinicians can easily interpret. Clinicians and patients need to understand scores and what changes over time might be expected. Importantly, PROMs scores should lead to specific actions. For example, physical functioning scores that are much more favourable than expected following surgery may prompt a revision of the recovery plan. Equally, mental health scores that are much lower than expected may start a conversation between clinician and patient, and result in a referral for further assistance.

The benefits from using PROMs may best be realised when outputs are interpreted in conjunction with other clinical information.

In quality improvement or policy development 

Processing of responses collected for non-clinical purposes is typically performed at an aggregate level rather than for each patient. The responses may undergo sophisticated analysis. For example, in benchmarking exercises, risk adjustment methodologies may be applied to ensure fair comparison between healthcare providers. PROMs outputs are typically interpreted in conjunction with other available information (for example, outputs from clinical measures).

PROMs can be interpreted by comparing a patient scores to their own score or by comparing with those from many other people. For example, scores from a health-related quality of life PROM may show that a patient has lower physical and social functioning than other people of the same age. Norms are often available for the general population and for people with specific health conditions.