The patient has, rightly, become a rising force in Life Sciences. Patient-centered care is about being responsive to patients’ needs and values, and this has been a guiding principle within pharmaceutical companies as well.

Often, drug (and biologic therapy!) developers are looking for ways to understand patient concerns, design clinical trials around them, and report these patient-centric findings. To do so, they utilize Patient Reported Outcomes (PROs). PROs manifest in two important ways: instruments and symptoms.

Instruments are questionnaires which capture PRO data used to measure treatment benefit or risk – this includes aspects of quality of life from general health to specific conditions (e.g., vision). Symptoms are just that – symptoms – but in this context they are reported by a patient (or caregiver) and usually affect one’s quality of life. For instance, there are voice issues associated with head-and-neck cancer, while brain cancer can be associated with debilitating headaches.

Within this area of research, people often ask themselves the following questions:

  • What is the landscape of symptoms or instruments, or even both?
  • Which instruments are associated with particular symptoms?
  • Where are the gaps? Do we need a new instrument?

Answering these questions requires a literature search – it’s a long and challenging process, potentially taking many months to truly consider all of the literature across time. Therefore, people use short-cuts to make the work more feasible (for instance, arbitrarily limiting the amount of literature considered such as reviewing only the “last 3 years”). This means they may miss things.

Instead, we’re introducing EvidPRO – a first-of-its-kind approach to PRO strategy that leverages the power of AI. We’ve written about the process before (though it has since evolved!), but the crux of it is that the AI surfaces symptoms and instruments related to a search (such as a disease state or drug class), all backed by the literature, and presents the results in intuitive and insightful visualizations.

Now, Evid users are no longer limited to the reading capacity of a human.  The powerful AI unlocks and unveils all-time PRO data reported on a subject, dramatically reducing the time-spent, from months to hours. As one user put it, “Rather than saying, I looked for PROs from 2015 to 2020, I was able to say I scanned all PROs within breast cancer. Period.”

Digging in a bit more… each question can now be answered…

What is the landscape of symptoms or instruments, or even both?

The first question anyone interrogating PROs should ask is “what has been done before?”

In other words, what is the landscape of symptoms that are most frequently reported by patients, and how are they are measured?

Consider brain tumors. This devastating disease is associated with severe cognitive symptoms combined with physical issues, and anxiety and depression about what the future holds. But what is the landscape of symptoms measured in this regard?

To answer this, the EvidPRO AI scours the published literature to unearth the most common patient-affecting symptoms.

As shown, the AI uncovers a number of cognitive issues (intelligence, vision, language, social, behavior), physical ones (headaches, activity, pain, fatigue, etc.), as well as psychological issues patients have while dealing with this disease (anxiety and depression).

EvidPRO unearthed both general and specific patient-reported symptoms associated with this disease, backed by nearly 1,500 published articles, all in an hour.

As one would expect, running EvidPRO for other areas yields different search-specific symptoms such as uncovering swallowing issues for head-and-neck cancers or cosmetic issues related to alopecia (as one deals with social anxiety over hair loss). In this way,  the AI reveals what symptoms are most important to patients.

Separately from the symptoms, researchers often want to know which PRO instruments have been most commonly used, within a given disease or therapy area. Essentially, “what has come before?”

In the example below, we see that for brain cancer, EvidPRO uncovered both general PROs, such as the SF (e.g., SF-36 and others), as well as some specific ones, such as the BRIEF questionnaire.

Understanding this landscape of previous work is important to determine if there are PRO measures that have precedent (which can be understood by reviewing which instruments have been used in later phase trials), to survey the possibilities, and further ensure that one is using the appropriate instrument.

Which instruments are associated with particular symptoms?

With knowledge of the symptoms and instruments at hand, it’s natural to ask about how they relate. EvidPRO can answer this question too. Specifically, one can ask – Given the symptoms I’m interested in, what PROs have been published around them? In other words, what PROMs are associated with these symptoms?

Sometimes, it’s obvious. For instance, for “vision” symptoms, one instrument is the VFQ-25.

Sometimes, it’s less obvious. For instance, a symptom of “headache” has been associated with QLQ-30 and SF-36. In other words, researchers associated QLQ-30 and SF-36 scores with headache, either directly (including it in the survey) or indirectly (measuring the effect of headache on the instrument).

In this way, EvidPRO helps answer the question and inform the PRO strategy. It allows one to figure out which instruments might be included in their clinical trial protocol, what associations have been made between core symptoms and instruments, how to conform studies to the current standard instruments, and even where there might be room for improvement.

Where are the gaps? Do we need a new instrument?

Finally, after reviewing the wealth of rich information EvidPRO provides, one may discover that there is not a validated instrument for a frequently reported symptom that’s been surfaced.

If nothing fits the bill, the researchers could propose their own tool to meet the need, and create the foundational version. Not only is this a research opportunity to quantify something that matters to patients, it’s an opportunity to create a standard for those areas where a treatment might differentiate itself from the rest.


EvidPRO brings the power of AI to bear on an increasingly important area – Patient Reported Outcomes. So if any of the above questions, or any PRO questions in general, arise, please don’t hesitate to reach out.

Read Also