She is in her kitchen, the laptop is propped on a stack of cookbooks, and the assessment requires observing her gait. What the camera shows is a countertop and, intermittently, the top third of a doorway. The visit is scheduled for thirty minutes and eleven of them are gone.

Remote visits solved a genuine problem. They also introduced a set of new ones that sites are still working out, and pretending otherwise does not help anybody.

Justin Coffeen has run visits in orthopedic, oncology, and hepatology trials. He lives and works in Lakewood Ranch, Florida, and came into research from hands-on clinical practice, which shapes how he thinks about what a screen can and cannot carry.

Some assessments transfer and some only appear to

A symptom questionnaire administered over video is close to the same instrument administered in a room. A range of motion measurement is not, and neither is anything requiring palpation, a calibrated device, or a standardized starting position.

The risk is not that sites knowingly substitute an inferior assessment. It is that a remote version of a physical measure feels like the same assessment while producing data with different variance. Coffeen’s background performing Functional Capacity Evaluations makes him skeptical of measurements taken under conditions the evaluator does not control, because standardization is most of what makes those numbers comparable.

The reasonable path is to decide, protocol by protocol, which assessments are remote-eligible and to say so explicitly in the study’s operating procedures. Ambiguity here gets resolved case by case under time pressure, which is how variance enters a dataset without anyone choosing it.

The participant’s home is not a controlled setting

In a clinic, lighting is consistent, the floor is level, there are no other people in the room, and the participant is not managing a dog.

Every one of those variables moves in a home visit, and several of them affect what a participant will say. Someone who lives with family may answer questions about mood, pain, or adherence differently when those family members are within earshot. Coffeen asks directly whether the participant is somewhere they can speak freely, and he treats a hesitant answer as a reason to reschedule rather than proceed.

Some of this is manageable with preparation. A short call before the visit covering where the participant will sit, what the camera needs to show, and what they should have with them removes most of the friction. The rest is genuinely outside anyone’s control.

Technology failures are documentation events

A call that drops mid-assessment, a platform that will not load, or a participant whose connection cannot sustain video are all real occurrences with real consequences for the visit record.

They need to be documented as what they were, including what portion of the visit was completed and under what conditions. Sites that quietly reschedule and record only the completed visit lose the trail of why a window was missed. Coffeen documents the attempt, because the attempt is part of what happened.

There is also a participant burden question hiding in these failures. Someone who has had two calls fail has spent time and attention on visits that produced nothing, and that experience has a cost in willingness to continue. Coffeen treats repeated technical problems as a retention risk, not only a data one.

Access improved, and not evenly

Remote visits genuinely widen participation. Someone who cannot take a full day off work, or who lives two hours from the site, or who cannot drive, may be able to join a study that was previously out of reach.

That gain is real and it is not universal. Participants without reliable internet, without a private space, or without comfort using the platform can find a remote-first study harder to complete than an in-person one. Coffeen’s read is that remote visits are a tool that expands access for some people and narrows it for others, and a site that offers only one modality has made a choice about who its study is for.

The question was never whether remote visits are good. It is which visits they suit, for which participants, under what documentation. Sites that answer that specifically get the benefit. Sites that adopt the format wholesale inherit the variance.

BLOGGER POSTS