Redefining recovery in fracture care: what new outcome measures mean for your patients

What does it really mean for a fracture to be considered healed? For surgeons, a good outcome after fracture has long been defined by radiographs and complication rates. For patients, recovery is about improving pain, restoring function, and returning to daily life. As fracture care evolves, so too must the way we measure outcomes. The AO Trauma Clinical Priority Program in Fracture Outcome Measurement has been investigating how innovative and patient-centered tools can help bridge the gap between what clinicians define as success and what patients’ views are on their own recovery journeys. Here, principal investigator Marilyn Heng explains why we should reconsider how we define successful outcomes.

For most of medical history, what has been considered by members of the orthopedic community as the successful treatment of a fracture has been defined in rather narrow terms. Historically, surgeons have limited their definition of surgical “success” to a handful of long-established—but seldomly questioned—clinical indicators. However, there can be discrepancy between what these indicators point to and what patients actually experience. 

Of course, radiographic proof that a bone has indeed grown back together remains essential, as does the absence of infections and other complications. As a discipline however, we are realizing more and more that our traditional ways of measuring how well we are doing in our treatment of patients may be somewhat incomplete: they may not fully reflect a patient’s journey to recovery.

Why traditional fracture outcomes may miss what matters to patients

When a patient comes in to see their surgeon, their number one question is: “So, is my bone healed, doc?” In the past, our answer would have depended largely on very disease-centric outcome measurements. For example, even today, we primarily reference an x-ray image to say yes or no. But as any orthopedic surgeon will tell you, even though a fracture looks completely healed on x-ray, we can still get patients who report they are really doing well at all.

To patients, full recovery means much more than successful bone union—it is about how much pain they are in, how well they sleep, how quickly they are able to return to an active lifestyle, and how they feel as they go about their everyday lives. That is why there has been a shift in recent years toward more holistic, patient-centered outcome measures: How can we better understand what is important to our patients—and are we actually achieving their goals for their treatment?

This shift can in turn be seen as the result of more general cultural and ethical changes in western societies, which have led people to participate much more actively in many aspects of their lives. Medical care is one such aspect: for the longest time, medicine was a very paternalistic discipline—patients did what the doctor told them. But the ethics of providing medicine have evolved over the decades, and principles such as patient autonomy and informed consent have become increasingly important. Consequently, patients today are being included much more in the decision-making process.

The shift toward patient-centered outcome measures

The research we are conducting under the Clinical Priority Program (CPP) in Clinical Outcome Measurements in Fracture Treatment very much reflects this evolution. Funded by the AO Foundation via the AO Trauma Network, the CPP reflects the AO Foundation’s belief that patient-reported outcome measures have a key role to play in the long-term evolution of fracture care. However, not all outcome measures are equal. How a measuring method is developed determines its reliability, sensitivity and overall usefulness in terms of actual patient care. As an integral part of the CPP, which consists of six separate subprojects, we are trying to establish which outcome measures best reflect what both the patient and the physician are hoping to get out of the care.

Some of the measures we investigate are patient-reported, where they tell us how they are doing. Others are technology-based—objective outcome measurements obtained from the patient, but not reported by them. We are also examining the predictive potential of certain approaches: are there measurements that we can make early on in the course of treatment that can help us better project how a patient is going to do at the end of their treatment so that we might be able to tailor their rehabilitation protocols?

Ultimately, we are hoping to establish a set of meaningful patient-centric measurements and combine them with some of the more traditional outcome measurements that have been used in fracture care, such as x-ray.

Legacy outcome measures vs modern tools

One of the challenges regarding patient-reported measures in a medical context is the fact that the term is very broad, and even somewhat nebulous. It can mean any number of things, and there is a definite need for clarification and structure. This is especially true as there are many legacy measures, as we call them. These are instruments such as questionnaires and surveys that surgeons developed to give to their patients. The first legacy measures go as far back as the 1990s.

Many of these legacy measures are still in use today. However, in many cases, they were designed by practicing clinicians who, although there is no question about their orthopedic expertise and qualifications, may not always have applied the most rigorous scientific standards on how to design questionnaires and surveys (also known as “psychometrics”). Surgeons would simply formulate a series of questions they considered important and assign scores, often with a total score of 100 and then arbitrary score thresholds: zero to 20 would be poor for instance, 20 to 40 would be fair, and so on.

Thus, many legacy measures were developed without any kind of psychometric validation. In fact, when they were retrospectively assessed for criteria such as reliability, it regularly turned out that they did not perform very well from a psychometric point of view. In some cases, there were large ceiling or floor effects where a lot of patients achieve very high or very low scores respectively, making it difficult to differentiate between them.

The National Institutes of Health (NIH) began funding research in 2004 to establish a series of patient-centric measures called PROMIS® (Patient-Reported Outcomes Measurement Information System). It was the first organized push to not only ask patients more, but also to make it more objective, rigorous and structured and to ensure the results are valid by applying the science of psychometrics.

Making PROMIS® relevant across age, language, and region

As PROMIS® was originally developed in the United States, its mean values reflect those of the average American citizen. In fracture care however, for example hip fractures, patients are not necessarily representative of the average population—we generally have more elderly patients. And we also have to take into account that there are geographical differences. That is why one of our subprojects focused on translating various PROMIS® questionnaire items into German and Spanish, which were then validated. Another subproject, which was led by a group of colleagues from Charité – Berlin University Medicine, focused on the question of how age and sex impacted the mean scores of certain PROMIS® scores in different countries.

The idea is to make PROMIS® scores more relevant to the patient context: how can we ensure that the PROMIS® physical function score of, for example, a 76-year-old female in Germany with a hip fracture, is comparable to her peers as opposed to what we would assume to be a younger mean individual of the general American population? The Charité team went on to create a web app that allows users to enter their score, country, sex, and age. In return, it displays the mean scores of that specific group of patients, as opposed to just the scores for the average person in the United States.

Harnessing technology to track fracture recovery

An example for the technology-based measures we are investigating are self-contained insole that can be placed in patients’ shoes. Equipped with pressure sensors, a six-axis gyroscope, as well as a temperature sensor, the insoles continuously record data such as ground reaction forces, pressure distribution, and foot gait. The data can be stored and transmitted via a mobile app. The primary objective of this subproject is an analysis of the association between automatically collected data on patient activity and loading, radiographic outcome measures and patient-reported measures.

During fracture healing, patients do not see their doctors every single day. Instead, check-ups only take place every few weeks or even months. However, and this is by no means limited to fracture care, we know that there is what we call recall bias: often, when we ask them how they are doing, they will only tell us how they have been doing right around that moment. We do not necessarily get a clear idea of the four weeks preceding. The insoles may give us a better picture of what happens between visits to the doctor—can we see how many steps they were taking? How much weight were they putting on it? The data delivered by the insole sensors could potentially allow us to individualize a patient's weight-bearing status maybe earlier, or hold them back a little more than we normally would.

Helping surgeons adopt new outcome measures in practice

As part of the CPP, we are also addressing a critical issue that is not generally explored enough: why do the results of new research not translate into clinical practice as a matter of course? Part of the answer to this question is that people, and especially researchers and surgeons, tend to use what they know. To them, usability matters, as do existing methods—no matter the rigor of new research.

As a result, there is still wide variety of legacy measures currently in use. We have to face up to the fact that not everybody is going to switch over to PROMIS® right away—even though the research we are conducting on the CPP is providing solid evidence for the advantages of psychometrically sound measures over older approaches.

We are therefore trying a different approach: taking inspiration from a larger project named PROsetta Stone, one of our subprojects has been creating crosswalk tables to convert scores between different measures. It is an attempt to translate the results of legacy measures into PROMIS® scores that focus on the same domain of physical function.

The hope is that if we can make PROMIS® applicable in cases where somebody wishes to continue using a legacy measure, or in cases where someone has old data from a legacy era, then maybe it can be something like a soft push to maybe convince them to consider using PROMIS® instruments in the future. Yes, it is difficult to come up with things that get clinicians and physicians to change behaviors and practice, but to me, it is part of what researchers should strive for when it is needed.

Customized care, better recovery experiences

Redefining recovery is not about replacing surgical expertise or tried and tested approaches. We are not trying to supplant radiography and other proven outcome measures. What we want to know is whether we can complement the instruments we already have at our disposal with deeper patient insight and improved data to hopefully help us take better care of our patients. Ultimately, the goal of healthcare should be to get the right treatments to the right patients at the right time.  We believe that systematically improving our outcome measures will ultimately enable us to better predict recovery trajectories, tailor patient instructions, support shared decision-making, and help set more realistic goals.