How to Spot a Badly Designed Nudge

Most people tend to ask the question: How do you design a good nudge? But it may be even more important to consider how to identify a badly designed nudge.

Nudges already exist everywhere, especially in health care. Think about the way information is framed and delivered in electronic health records, patient portals, and conversations between clinicians and patients. Poorly designed nudges may not only adversely impact our decision-making but it may also distract us from being able to focus attention on the well-designed nudges.

When considering interventions targeted to clinicians to improve the delivery of health care, here are 4 red flags that may tip you off to a badly designed nudge.

1. The nudge is simply a reminder.

Perhaps the most common missed opportunity is designing a nudge that acts simply as a reminder. This design focuses on raising awareness to an issue and assumes the main barrier is forgetfulness. But that often is not the case.

Take for the example, a primary care practice that is trying to improve cancer screening rates. They could decide to trigger an alert in the electronic health record that reminds clinicians to discuss cancer screening with their patient during a clinic visit. Surely this is well intended, but it can quickly lead to alert fatigue – whereby a clinician gets tired of these alerts and starts ignoring not only these but other alerts as well.

Primary care clinicians already know that cancer screening is important. A well-designed nudge could focus more on driving motivation and reducing the effort needed to take action. Read on for examples of nudges that do just that.

2. The nudge does not fit well within existing workflow.

One of the most common reasons a nudge misses the mark is because it wasn’t design with the clinician’s workflow in mind. This often occurs when there are multiple pathways to complete an action and the wrong one is selected.

For example, let’s say you want to target antibiotic prescribing. Many health systems have created ordersets in the electronic health record for types of drugs (e.g. antivirals) or conditions (e.g. pneumonia). You could spend a lot of time changing the design of these ordersets to later find out that most clinicians never use them. Instead they just search for the name of the drug they have in mind.

When designing a nudge, its important to work directly with the end users of the intervention (e.g. clinicians) to learn about their current workflow and how to design your intervention to fit within it. EHRs can provide important information on which pathways are most commonly used to place orders.

3. The nudge is delivered to a group that is different from the group implementing it.

One of the keys to implementing an effective nudge is to ensure stakeholder alignment. This is one of the most critical components of designing a nudge. Unfortunately, it is also one of the most overlooked. An example that highlights this is when one group comes up with a nudge that is targeted to a completely separate group without getting buy-in and support from that group. For example, if cardiology came up with an idea to nudge surgeons to place an order for something right after surgery but didn’t involve the surgeons in the implementation process.

These nudges could potentially be appropriate but they are unlikely to be well-received or fit within clinical workflow if the group receiving them wasn’t involved in the design.

4. The nudge increases rather than reduces work.

Clinicians are busy. Nudges that require a lot of time and effort are unlikely to be successful.

In a clinical trial, we tested whether providing clinicians with a list of their patients who met national guidelines for a statin medication to reduce the risk of heart disease would increase their prescribing of these drugs. While prescribing rates did increase from 2% in the control group to 6-8% in the intervention groups, they were far from optimal. After the study ended, we asked clinicians why it didn’t’ work. Their response: these lists are great but they can be really long and we don’t have time in our day allocated to review and act upon them. We are busy in clinic seeing patient after patient.

The most effective nudges reduce work for the recipient. This can be done by automating what were originally manual processes. For example, in a cardiology project focused on patients discharged from the hospital after a heart attack, we were able to increase referral to cardiac rehabilitation from 15% to 80% by transforming a manual process tasked to busy clinicians to an automated one that identified eligible patients and templated all the necessary paperwork.

It can also be done by making the right thing the easy thing. Generic medications work exactly the same as their more expensive brand name drugs. By changing prescription settings in the electronic health records to default to generics unless the clinician opted out, generic prescribing rates increased from 75% to 98%, literally overnight. These effects were sustained and over 2.5 years it saved more than $32 million.


Badly designed nudges are more common than they should be in health care. Avoid these common pitfalls and make sure your nudge is on target.

_Have a question or want to reach out? _Click here_ to send Dr. Mitesh Patel a message. _