How many PROMIS items do I need to calculate a PROPr score for them?
We suggest a minimum of 2 items per domain. With the 7 domains of PROPr, this means a minimum of 14 questions. If you would like to take this approach, we recommend using the 14 items used in PROPr development (see page 17 of the whitepaper for items).
There are three common ways to collect the data necessary for PROPr that are easy to score:
- Administer computer adaptive tests (CATs) for the 7 PROPr domains. If the CATs take 4 or 5 questions to complete per domain, this would be around 32 questions.
- Administer the PROMIS-29+2 Profile v2.1. This is the PROMIS-29 Profile v2.1 and two items from Cognitive Function – Abilities v2.0. This is 31 questions.
- Administer the PROMIS-29 Profile v2.1 and a PROMIS Cognitive Function – Abilities or PROMIS Cognitive Function short form (e.g., 4a). This is 33 questions.
If you would like to create a custom short form, please see this page.
Do I need a score on all 7 PROMIS domains in PROPr to get a PROPr score?
The best way to get a PROPr score is to have all 7 domains. However, because the PROMIS Profile instruments (such as the PROMIS-29) had wide use before PROPr’s development, we have developed a method to predict a PROPr score using only Profile domains.
I have PROMIS-29 data. Can I use PROPr?
Yes.
I have PROMIS-Global/PROMIS-10 data. Can I use PROPr?
You can map to PROPr using this algorithm.
Can I use PROPr to inform individual-level decisions?
No. PROPr is constructed using societal preferences; PROPr is meant to represent aggregated preferences. An individual may have different health preferences than the aggregated health preferences for PROPr. This is much like how an individual may have a different candidate preference than the results of an election.
Why is PROPr based on preferences elicited using the standard gamble?
The standard gamble (SG) is a well developed method with strong normative grounding in multiattribute utility theory. It also easily allows for “decomposed” methodology in which each health domain is valued individually and then they are combined into a final model.
One important feature of decomposed methodology is the ability to test if the final model is additive or multiplicative. So far as we know, the additive model has been rejected every time it has been tested in favor of a multiplicative model in health status utility modeling. This was also true for PROPr, so PROPr has a multiplicative model.
Why is PROPr based on preferences collected from a community sample? Can I use it if my PROMIS scores are from patients?
PROPr uses community preferences (a representative sample of the adult US population) so that it can be used to make societal decisions such as decisions about resource allocation.
Yes, you can use PROPr on data from any population. For example, let’s say that you are interested in patients with disease A. You can collected PROMIS data from a sample of patients with disease A which provides information about their health status (e.g., their physical functioning status). You can then calculate PROPr from these data to determine how the general population values the health status of the sample.
Why should I use PROPr instead of legacy utility measures?
The most widely used generic preference-based measures include the EuroQol-5D (EQ-5D), Health Utilities Index, SF-6D, and the Quality of Well-being Index. Each, however, has some limitations: (1) large proportions of the respondents scoring at the very top or very bottom of the scale (i.e., ceiling effects in the very healthy or floor effects in the very ill), (2) imprecise measurement for individuals, (3) poorly-worded questions such as those that combine concepts (double-barreled items), and (4) lack of coverage of the full range of health. These limitations arise from the descriptions of health used in these measures and not the method of scoring.
PROMIS, being constructed using item response theory, addresses several limitations of the existing generic preference measures including: (1) capturing a wider range of each health domain, (2) measuring individual health status with greater precision, and (3) using rigorously designed and tested questions.
PROPr is the first preference-based summary score to link directly preference-based functions to health domains as measured by item response theory. As such, PROPr gains many of the advantages of PROMIS, including flexible administration of items from the health domains used to construct PROPr.
The use of PROMIS/PROPr allows the same items to be used as a health status measure and as a health utility measure.
Why is the mean PROPr score lower than those of the legacy utility measures?
We have shown that PROPr scores are lower than EuroQol-5D-5L, Health Utilities Index Mark 2, and Health Utilties Index Mark 3 scores in samples where they were coadministered. We expected PROPr scores to be lower than these legacy measures because the best health state described in PROPr is significantly better than the best health state described in legacy measures. For example, the best physical functioning in PROPr is “able to dress yourself, including tying shoelaces and buttoning up your clothes without any difficulty and able to run 100 yards (100 m) without any difficulty.” In contrast, the best physical functioning in the EuroQol-5D-5L is “I have no problems walking,” and in the Health Utilties Index Mark 3 , “I have full use of two hands and ten fingers and I am able to walk around the neighbourhood without difficulty, and without walking equipment.”
The increase in descriptive space in PROPr “raises the bar” to reach a best-health score of 1.0. This will both reduce ceiling effects in the general population and substantially lower scores when PROPr is compared to legacy measures.
What it this minimally important difference for PROPr?
The minimally important difference for PROPr has not been formally evaluated, but we currently recommend using 0.04. Most preference-based measures have minimally important difference thresholds between 0.03 and 0.05. Current work-in-progress suggests a minimally important difference of 0.04 would be appropriate, although a conservative estimate of 0.08 (half of a standard deviation in the PROPr dataset) could also be used.
Has PROPr been validated?
There is currently cross-sectional validation of PROPr against many of the legacy instruments looking at demographics and chronic health conditions. There is also a cross-sectional validation of PROPr against many of the legacy instruments focused on social determinants of health. Other validation studies, including longitudinal studies, are currently in process.