Your medical records don’t contain all the information you think they do.
If there’s one thing I learned from spending nearly 14 years of my career reading medical records to make Social Security Disability decisions, it’s that your medical records don’t contain all the information you think they do.
When it comes to electronic medical records, how much of the data problem is because of poorly designed software due to poor requirements elicitation? How much did the creators of the software think about the end consumer of the medical records—the folks reading those records later, the patients, the Disability Examiners, or other medical professionals reading those records?
When Medical Clinics and Hospitals select an Electronic Health Records (EHR) system to implement in their facility, do they think about the consumers of the records that they’ll generate?
Digital Records Are Better - Right?
Disability claimants would routinely tell us that all the information we needed was in their records. Just look at the records, they’d say. But, often, all the information we needed was not in the records.
I saw all kinds of medical records: hand-written records, transcribed records, records from other countries, clinic records, hospital records, therapy records, worker’s compensation records, military records, VA records, etc., and I got to witness as more and more doctors and hospitals made the shift from paper records to electronic records.
I came to the decision that as a consumer of medical records, electronic medical records are a mixed blessing. There’s that joke that doctors have indecipherable handwriting. It’s often true. Electronic records are great because you don’t have to spend time trying to read writing that’s haphazardly scrawled across the page, while keeping up with all the medical terminology and abbreviations.
But you know what’s great about hand-written medical records? You can trust that the doctor actually wrote those notes about the patient for that day. You know they didn’t copy and paste those notes from the previous visit.
Signs of Trouble
The exact type of information and detail a Disability Examiner is hoping to gather from medical records might be fairly different from a patient or another medical provider, but everyone consuming a record expects to get true and accurate statements regarding a patient’s history, current exam, function, plan, and medications. The problem I saw with so many medical records was one of trust. Do I, as a reader of this medical record, trust that this information is accurate? Do I trust that the doctor actually did the exam that the record says they did?
And you may ask, WHY didn’t I trust the medical record I was reading?
There were several common themes:
Evaluating Medical Record Quality
- Obvious copy and pastes. The same history and same plan would be listed medical visit after medical visit. If the doctor copied and pasted that much of the record, how much of the other parts of the exam are copied and pasted too? Was everything during the exam truly the same as the last visit?
- Long current medication lists with frequent duplicates, that were often contradicted by what the body of the medical report said. I understand that it is helpful to have a record of the history of medications prescribed to the patient, but when that list seems to be jumbled with the current medications, you can’t trust the information on the medications list.
- Unclear or lazy use of notations—particularly around the use normal (nl), or Within Normal Limits (WNL). For example, if someone’s gait is normal, the doctor could describe their gait as nl or WNL, but I routinely saw medical records describe a patient’s gait as WNL when other records indicated that it wasn’t normal. Was the doctor just clicking through the EHR system too fast? Or did the doctor mean that the patient’s gait is within normal limits for the patient (which may not be WNL for the normal population)? In disability determination, if we had a discrepancy like this in the medical records, especially if someone was alleging disability related to difficulty walking, this created a problem.
- Records showing that a doctor did an exam that they were unlikely to have performed. For example, orthopedic records showing that a patient’s eyes were PERRLA (Pupils Equal, Round, and Reactive to Light and Accommodation), or their lungs were CTAB (Clear To Auscultation Bilaterally). Did the orthopedist truly examine this patient’s eyes, or listen to their breath sounds? Records also routinely include a patient’s gait, strength, and reflexes (as normal). How often are gait, strength, and reflexes really examined by a medical provider? I know that I’m generally sitting the whole time I see a doctor, and I can’t remember the last time my strength or reflexes were assessed. Do these portions of the exam show up and autofill in the medical record because of the EHR system the medical provider is using?
- Records that refer to the patient by the wrong gender. This makes you question if the records really belong to the patient or were accidentally input into the wrong electronic medical record.
Consider ALL the Stakeholders
A medical provider’s decision about what EHR system to implement might look at cost, ability to integrate to existing systems the provider uses, ease of use, the features, and the ability to customize the system.
An EHR software provider will likely include features related to recording and managing the patient’s medical history, exam findings, medications, care plan, ordering tests and labs, referrals, handling follow-up information, scheduling, handling medical billing and ICD-10 coding, insurance, payments, and may include patient portals and features related to communicating with patients.
Do the medical providers consider all the consumers of the medical records that they are creating in the EHR software selection process? Would a better requirements elicitation effort in the EHR software creation process solve for these data integrity and trust issues? I would like to think so. What do you think? Let us know! We’d love to hear from you.