August 13, 2019

Ways to search for a patient in the EHR

INTRO / SUMMARY
Watch this article as a video

The previous article emphasizes that the primary purpose in an electronic health record of the HOME screen is to select which PATIENT FILE to open. A well-built HOME section will display the next PATIENT FILE the clinician needs.

This post explores some of the different ways it is possible to display the next PATIENT FILE needed.

The first consideration asks the designer to decide what type of lookup strategy is needed.

Does the lookup need to happen if:

A. The patient may be or may not be physically present at the time of the lookup?
   
1a. Scan barcode or QR code
    2a. Manually search for patient by number
    3a. Manually search for a patient by name
    4a. Select a patient from a patient queue during an active clinic

B. The patient is likely not physically present at the time of the lookup?
   
5b. Select a patient from a patient list

C. The patient is always physically present at the time of the lookup?
   
6c. Search by biometric

In short: it is likely best to first add to the EHR search by number (and name), and then add additional options later.

SECTION A: The patient may or may not be physically present at the time of lookup:

1a. SCAN barcode or QR code

The quickest way to identify the right patient file in almost any circumstance is by scanning a barcode or QR code. This lookup strategy also works if the patient is not physically present, provided their paperwork has a barcode / QR code attached.

Advantage: fast & produces a reliable 1:1 match; no typing needed

Disadvantage: an ID barcode or QR code for the patient is necessary, as is a device to scan this. The scanning device can be either the built-in camera on the phone or an inexpensive USB/Bluetooth barcode reader. A potential disadvantage is that regular use of the phone’s camera to scan patient IDs may decrease the device’s battery life.  However, I'm not sure the real-life impact of this.

2a. Manually SEARCH for the patient BY NUMBER

Searching by patient number is easily the most robust method applicable across a wide variety of scenarios, devices, and care organizations.

Searching by a patient's number works in almost all situations, with organizations of any size, with high reliability, on any device type. This is why it is my default 'go-to' search method.

Advantage: produces reliable 1:1 search results. Is fast to enter because there are a limited number of digits, and ideally they are all numbers. By default, the numerical keypad on screen can appear, and this is fast to operate on multiple device types. A fast ‘check’ to ensure the numerical lookup was correct, is ensuring the patient name is correct.

Disadvantage: The patient may not have a unique identifying number. (Though this seems rare, you should be able to find a national ID number, a locally assigned number given out by your organization, or a phone number).

3a. Manually SEARCH for the patient by NAME

This also works but can be more tricky. Patients may swap the order of their names, go by different names, and have similar names.   The person entering the name may also not know how to spell it.

A name matching service/master patient list service is required to help overcome this to find similar-sounding names, name variants, and misspelled names.

There is also a high chance that the search result will yield multiple patients with the same name. The user will have to be very careful to ensure that they are selecting the correct match. In order to do this, they must verify other information such as: other names, relatives, contact information, date of birth, or clinical issues and last appointment date).

A variant of this approach, for the patient that cannot be identified by number or name, is to manually search for them by other information: such as a relative, an email address, a household or town. The effectiveness of this is based on how large of a catchment the system searches.

Advantage: works when the patient doesn’t have a unique identifying number on them. In general, I’d consider this a ‘backup’ strategy to search for patients.

Disadvantage: The problem with this approach over searching by number, is it requires typing more characters into the screen, and has a higher chance of not producing a 1:1 result. This means more time to sort through search results, and a higher risk of opening the wrong patient file.

Part of the advantage of searching by number, is that the confirmation step the numerical search was correct, is that it matches the patient’s name. However, when one searches first by name, there is not a quick default second step to confirm that the correct patient has been identified.

Note: this strategy may require extra consideration if the organization has multiple identifiers that have the same number of digits. In such a case the same number may return two different patients - one with each identifier. In such cases, it is critical that the user knows what number they are searching by, and what number is being returned on the patient result.

4a. Select a patient in a PATIENT QUEUE during an active clinic.

Selecting patients from a patient queue is often an excellent method for clinicians working within an active clinic. However, it may not be the best search strategy for patients arriving at the front desk. Let me explain.

If a clinic has 40 or 60 patients scheduled for the day (or half-day), it likely is faster to enter a patient’s 8 digit code, than to review a list of 50 patients by name. The effectiveness of browsing that list depends on how it is sorted (eg. which first name did the patient provide, or which last name was given). Once found, a second confirmation step is still required after the patient has been identified on the patient queue because there is a high risk of identical patient names.

However, the patient queue may be very helpful for providers pulling up a PATIENT FILE after the patient has already checked into the clinic. In this case, the number of patient’s in the queue only reflects the number of active patients in the clinic. This is a small manageable number. The queue also provides additional information, such as the next patient to see and information that helps triage who to take in next.

If the goal of the HOME screen is to alert the user ‘who they should see next’, the patient queue is often the right way to do this.

Advantage: a patient queue provides a quick way to identify the next patient file in a clinic once the patient already has checked in.

Disadvantage: searching a long patient queue may be slower to check in a patient than searching for their name.

SECTION B: The patient is likely not present at the time of lookup

5b. Select a patient from a PATIENT LIST.

A patient list is useful in many situations. To differentiate a patient list from a patient queue, I use the term patient queue to describe lists of patients within an active care clinic (or care setting). This suggests a clear hierarchy and triage of arrival and reflects that patients are at different stages of their appointment and care. It also assumes the patient is physically available during the lookup process.

The patient list is also a list of patients but does not have to be related to an active clinic.

I think of the patient list as the higher-level idea, and the patient queue as a specialized subtype of the patient list.

A patient list may be used, for instance, to

  • follow up on patients with bloodwork results
  • contact patients who miss an appointment from a clinic
  • follow up newly referred patients
  • see patients on a ward by location

In each example, the patient list likely self-generates based on a set of list criteria (eg. bloodwork result needs followup, missed X clinic appointment, X days ago).

The patient list provides a very effective mechanism for the clinician to be able to see what work they need to do. Open the required files. And after dealing with the issue at hand, remove the patient off the active list. This workflow is similar to working one's way through an email inbox or to-do list.

Advantage: provides a very clear work list for the clinician to use for their job. Makes routine workflows and tasks easy.

Disadvantage:  requires that the search parameters are reliable in automatically generating the list.

SECTION C: The patient is always physically present at the time of lookup

6c. SEARCH by BIOMETRIC

Some consider this the “best” method to search for a patient. But it has been listed as the last. The use of a biometric, such as a fingerprint scanner sounds very appealing as a way to retrieve a patient’s file.

However, this requires that physical hardware for the biometric (such as fingerprint reader, iris scanner, etc), is in place. When running an organization with multiple devices (mobile, tablet, and desktop), across multiple clinicians, across multiple locations, this increases that chance of a hardware malfunction and additional cost.

The other problem of biometrics is false-negative results. For instance, from a patient who's fingerprint change over time, or if the device is operated by someone who doesn't know how to troubleshoot it.

Advancements in biometrics will enable accurate recognition of patients using only a device’s camera. When this is readily available, it is a very appealing method of identifying the desired patient file when the patient is physically present at the time of lookup.

Don’t conflate patient lookup and patient verification:

Pulling up the correct patient file is one thing. Checking that the patient in front of you is who they say they are (ie. verifying identify) is a different thing. This post focuses on pulling up the correct patient file, based on who the person claims to be. A subsequent post will comment on patient identity verification strategies.

A word on identical patient names:

Some may suggest this posts over emphasis the risk of identical patient names, and that my strong aversion to using this as the first search parameter is over-exaggerated. From my own experience, on a 20 to 30 bed medicine unit, it is quite common to have at least one or two or even three patients on that unit who share the same name as someone else on that unit. Imagine, the chance that a community healthcare worker who cares for a few hundred patients, or a clinic with thousands of patients will encounter similar patient files.

Do you know the patient?

This article discusses typical EHR environments, where (a) those who use the system do not know each patient by name, (b) where those patients on a list may change frequently, (c) there is potential for a large number of patients within the catchment search area. This is very different from other types of care environments, such as a community health worker, who knows each patients very well by name, and who’s list of patients does not change dramatically, and who’s total list length is of a confined amount. In such circumstances, the use of a patient list combined with a filter by name strategy may be most optimal for such settings.

Subscribe for updates

Get an email update weekly, monthly, or quarterly.
Subscribe