To Survey or Not To Survey

One thing that has really helped the universe of patient experience has been surveying patients.  One thing that has really hindered the universe of patient experience has been surveying patients.  It has helped because it has created a standard measuring stick that allows different hospitals, units, clinics, and departments to compare themselves with each other.  Even before standardized surveys like the suite of CAHPS surveys, the ability to compare and contrast with other hospitals in a system or units in a hospital helped identify hot and cold spots.  Even if folks wanted to argue that a patient’s experience in OB and in oncology was different with different patients with different care plans with different expectations, having a common language at least allowed for a start to those conversations.

At the same time, though, this measurement created a coin of the realm that would reduce all PX conversations down to a question of percentages and a comparison to benchmarks.  First, it effectively stripped the ‘experience’ from patient experience and replaced it with a trackable number upon which beatings could be delivered.  Second, and in my opinion more damaging, it effectively removed a large population of patient-facing folks in the hospital from any responsibility in patient experience.  After all, if there isn’t a food question on the survey, how can you hold food service accountable?  If there is no way to ask questions about lab, imaging, or therapy (or even identify patients who got those experiences) how can they have a seat at the PX table?  If a hospital does not discharge out of ICU, but instead discharges them from a step-down unit, how can the nurses and doctors in intensive care talk about the experience?

So, the very act of providing a common language also dumbed down the conversation.  I remember being at a conference fifteen years ago where the presenter was talking about the over-surveying of patients and they weren’t even talking about post-discharge surveys.  They were talking about every department reaching out and surveying patients or staff on a host of different things.  I didn’t believe that this was happening, until I started talking with clients and discovered that they might have eight to ten other surveys in-play other than the one I was helping them with.  I further discovered that there were even more departments that would have liked to run a survey but couldn’t figure out how to manage it. 

Outside of the fact that they felt like they needed something to put on a dashboard, some departments, like lab and pharmacy, might have to survey to check a box with their accrediting body.  The last time I read the accreditation rules for lab, for example, Commission on Lab Accreditation (COLA), College of American Pathologists (CAP) and Joint Commission, all required some form of experience measure.  These survey requirements were vague, but present.  So, while the specifics could be massaged, the need was real. 

One might imagine that the simplest solution here would be to simply add a question to an existing survey.  The challenge here is that, like so many things, this is simpler said than done.  Imagine that you wanted to ask your emergency department patients about an imaging element to their ED stay.  You have three options.

1. Ask everyone about imaging, and deal with irritating a large portion of your patients who did not have an imaging experience as part of their stay. 
2. Ask a gate-keeper question (Did you have an imaging experience as part of your ED visit?) and only ask imaging questions of those who said yes.  This requires a patient to know what ‘imaging’ means and remember if they had one.  It also adds a question on the survey for every patient in an environment where people are skittish when a survey gets longer than eight or nine questions.
3. Add a flag denoting imaging experience to the data that is sent to the vendor.  It is the most efficient way to target the patient population.  It is accurate and it doesn’t add an extra gate-keeper question to the survey that everyone must answer.  It does require your data folks to code and add the flag to the data which is either easy or impossible depending on how easy extraction from your EHR is, or how overworked your data team is.  It also does add a question or two to the survey’s overall length, even if it is asked only of a subpopulation.  And if you only add one question, I will guarantee you will add the wrong question.¹

Not only is survey and data architecture an issue here, if imaging wants to add a question to the ED survey, there are a lot of cooks in that kitchen.  They will have to appeal to the ED governing committee and the patient experience governing committee.  Depending on how byzantine the structure is, they might have to appeal to the clinical team as well and prove to finance that this won’t change the cost of the survey.

Or they can just write a few questions on a postcard or QR code and distribute it out to patients who hoof themselves into their field of view. 

So, lab can do it, imaging can do it, food service can do it.  Any contracted third-party service can do it as well.  So if you contract out your valet parking, or gift shop, or anesthesiology, they will also have a survey.  Heck, increasingly hospitals are executing a scheduling survey, asking patients to stay on the line after scheduling an appointment to take a brief survey.  It doesn’t take long for a patient to get handed a bunch of surveys.  The impact that this has on the patient is a topic I will save for another time.  God forbid I write another two-part essay.

The impact on the organization is multifold and generally not positive.  Yes, food service or housekeeping can put a number on their huddle board, and they have an easy win when a senior leader asks about their service action plan.  “We are tracking a number!”  But what actual value does that number have?  In the various organizations where I have seen it reported, it has very little value.

• The surveys are poorly constructed.  They have poorly worded questions with confusing scales.  They are written from the perspective of the clinical staff, so often they focus on issues patients may not care about, or, use words that patients don’t understand. 
• The surveys are poorly executed.  Many surveyed areas, like infusion, dialysis, or lab, are often used by patients who will have multiple encounters a month.  If not careful, staff will survey some patients multiple times and others not at all. 
• Ther data is not processed or not processed well.  People often don’t realize that there is work in turning hand-entered data in Excel or collected on an on-line surveying tool into a useful data display.  They may not know how to run a pivot table or crosstab to differentiate between multiple labs or trend by months.
• The data is seldom used.  It may generate a number in a box, but it is not being reviewed or discussed or used to change process or improve scores.  In fact, often you will hear people talk about the data as if it is completely removed from the actual experience.  “I don’t know why they said we didn’t introduce ourselves; we always introduce ourselves!”  One of my mottos is that data you are not using is not worth collecting.  You are spending money and time or money disguised at time running a survey.  You are annoying your patients by giving them yet one more survey.  You are taking more time to tabulate results.  And after all of this, you are not using the data to drive improvement?  Do not be surprised if your scores drop, since the only thing worse than not asking someone how they feel, is asking them and not really caring what they say.

Organizationally, it has a couple of other impacts.

• It becomes an excuse.  If a department cannot figure out how to quantify their patient/employee experience, or, they cannot get system support to develop a survey tool, they have a built-in excuse.  “We cannot address our PX because we don’t have a way to measure our PX.”  Once this excuse accepted by leadership, getting a department to create a service action plan is like pulling teeth.  The department can say that they don’t know what to do, or say that they are doing something, but cannot prove it is working.  They can conflate not having data with not having an impact on patient experience. 
• It changes how patients view a health care experience.  Most efforts in PX are focused on building a themed teamwork-oriented care experience.  So, an ED-admit to ICU followed by discharge from a stepdown unit to a post-discharge follow-up appointment with their primary care doctor was to feel like one seamless experience.  Somewhere right now in a hospital there is a team eating stale donuts and drinking marginal coffee while journey-mapping a patient’s travels through their care site so they can create a unique and branded experience.  But by breaking the experience down into distinct pieces, starting with a greeting and ending with a survey, creates a very episodic and transactional experience.  Quite the opposite of the seamless one everyone strives for.

So what do we do if we cannot survey our patients?  I will assume that you are asking this question honestly and not passive-aggressively.  Here are two sources to inform process improvement strategies and measure success.

Your own experience.  It is your patient’s first day; it is not your first day.  Reflect upon your encounters with patients in your space.  Ask your staff to think about their interactions in the care space.

• When a patient is complimentary, what are they singling out?  Is it our friendliness, our speed, our attention to detail, our teamwork? 
• When a patient is critical, what are they saying?  Is it a personality issue, a structural issue, a timeliness issue?  Is it universal, or limited to a provider, a time of day, or a type of patient?

And then you build action plans to do more of the former and less of the latter.  Even if it is something that feels outside of your control (like scheduling, the parking lot, office hours, etc.) you can still build an action plan around effective communication on that issue.²

Talk to your patients.  There are plenty of opportunities to chat with a patient while care is being administered.  I have mentioned in the past how effective my primary care doctor is at that.  You don’t have to quiz them, just open the door for conversation and patients will tell you what they see and what they want.  In fact, if you know what drives expectations, you can ask directly about that, getting the patient’s perspective.  Once you have made changes, you can even ask if they have noticed any difference.  “We restructured registration staffing for our busy times, did that seem faster to you today?” 

Conducting surveys is an efficient way to gather data.  But it can become a crutch.  It can actually make staff less likely to engage with patients, because they figure ‘that is what the survey is for,’ or they are concerned that they may violate survey rules.³ Homegrown surveys can be helpful and can satisfy an accreditation requirement, but like poetry, they are too easily done poorly.  Survey data can be a very valuable tool to start an improvement strategy, but if “do a survey” is your improvement strategy, you will fail.

¹Most people who want to add a single question will focus on an overall question, such as, “How would you rate your overall experience with imaging?” which is a good question to fill a box on a dashboard, but a crappy question to ask if you want to do process improvement.  If a patient gives you a 7 on the 0-10 scale of an overall imaging rating question, what do you do with that?  How do you action-plan around that?  There are a lot of better single-questions to add, but I will save that discussion for another time.

²As always I have a lot more to say about that.  One of the great aggravations I have is when someone says, “But I cannot control that!”  True, but you can control how you talk about that with your patients.

³CMS does have rules about CAHPS execution, including not reviewing the questions before the patient gets a survey.  So a nurse cannot ask a patient, “hey, while I have you, if you were to rate us on a scale from 0-10, what would you give us?”  This rules, like so much else, are vague, which means that there are some who don’t know you can’t ask patients CAHPS questions and others who think you cannot talk to a patient about anything related to experience.