Around the Web: Is This Data Reliable?

By Alessandra Suuberg, Decency LLC

From recent headlines to recent podcasts—and from traditional medicine to the world of consumer gadgets—here are some items on my mind with respect to the limitations of healthcare assessments and records.

In the News: Headphones that Test Your Hearing

I sensed a recurring theme this week, checking my inbox and seeing that a new product will join the ranks of consumer gadgets with healthcare functions.

The coverage in question described hearing-health features newly added to a popular brand of consumer headphones.

These new capabilities will take the headphones beyond their classic purposes, e.g., listening to music, and “allow users to test their hearing” and use the “headphones as hearing aids for milder forms of hearing loss.”

The coverage noted praise from audiologists and others, as “[m]any people with hearing impairment never get tested or seek treatment.” However, the same news outlet also predicted possible “confusion about [the headphone maker’s] lofty marketing and how doctors and users are meant to interpret it.”

It seems that when health-enhanced gadgets like these hit the market, questions tend to follow about their capabilities and limitations—and how they should fit into the world of traditional medical care.[1]

It will be interesting to see how this new product shakes up the healthcare world, and what kinds of studies and conversations it sparks among users and healthcare providers.

In the Podverse: Accuracy of Blood Pressure Readings

Of course, questions about reliability are not limited to the world of new consumer gadgets.

This past week, one of my favorite nursing education podcasts[2] touched on sources of inaccuracy in blood pressure measurements.

Specifically, in the context of an episode on hypertensive crisis, the podcast ran through some of the subtle factors that can lead to a falsely elevated blood pressure reading.

The cited examples included:

  • a cuff that is too small (potential impact: 2-10 mmHg);

  • the patient sitting without their back supported (6 mmHg);

  • crossed legs (5-8 mmHg);

  • the patient’s arm dangling, unsupported (10-12 mmHg);

  • the patient talking during assessment (up to 10 mmHg);

  • the patient actively listening (up to 10 mmHg);

  • a measurement taken over clothing (varies widely, up to 50 mmHg); and

  • a full bladder (up to 15 mmHg).

How Reliable Are Medical Records, in General?

If something as basic as a blood pressure measurement is so easily swayed, how accurate is the data in medical records generally?

The question of medical record accuracy was addressed earlier this year by one survey reporting that 45% of consumers had found inaccuracies in their records, though 67% said “errors ha[d]n’t affected treatment.” 

For me, this raises a few questions:

1. What are the most common inaccuracies in medical records?

The survey data noted inaccuracies with respect to “personal information (33%), allergies or aversions (23%), and treatment history (22%).”

In addition to these, I would be curious to know how often, e.g., falsely elevated blood pressure readings make their way into records.

2. How often do errors occur?

In other words, for any given patient’s medical record(s), what percentage contains inaccurate data?

3. How and how often is the information in medical records corrected?

According to the survey results, “[m]ore than half of Generation Z” and millennial respondents “ha[d] had to correct or ask for clarification on their health records.” 

4. How often do these errors have consequences for the patient?

According to the survey data, 67% of respondents did “not believe inaccuracies impact[ed] their . . . treatment,” while “33% [of respondents] cite[d] wasted time, inaccurate treatment, and prescription issues as significant problems.”

Finally, the survey also brings attention to artificial intelligence and its potential impact on record accuracy. The survey results raised a possibility of using AI for “detection of incomplete information and duplicate records.”

This raises the question:

5. Will artificial intelligence make medical records more or less accurate?

This seems to be another area ripe for interesting future studies and conversations.

[1] Past studies have examined the accuracy of consumer devices with respect to, e.g., measurement of heart rate, aerobic capacity, blood oxygen saturation, and sleep time.

[2] Decency LLC has no relationship to this podcast.

Disclaimer: The information and opinions on this site do not include legal advice or the advice of a licensed healthcare provider.