The challenges in estimating health system response to waiting time measurement systems

Posted on May 27, 2012 by aviyer2010

An article in the Wall Street Journal (May 25, 2012) titled “Long Medical Waits prove hard to cure” describes different waiting time measurements and the consequent dysfunctional system response. Measuring the waiting time experienced by patients served or focusing on service within 48 hours creates an incentive to prioritize more recent patients to improve performance. A focus on the number waiting to see a specialist creates an incentive to wait to add patients to the queue. A Veterans Administration report claims that while records suggested that 95 % of patients seeking mental health waited less than 14 days, the actual % was 64 %, if one considered the total wait time. But perceptions of patient waiting time can also be erroneous. How should the system be measured to decrease the incentive to create unintended consequences ? Should the referral process be tracked across doctors and clinics to ensure end-to-end measurements ? How should the patient’s perspective be ensured in these measurements ? Could incentives be designed that wille ensure prompt service while maintaining quality ?

About aviyer2010

Professor

View all posts by aviyer2010 →

This entry was posted in Operations Management, Service Operations, Supply Chain Issues and tagged Consumers, Coordination, emb, healthcare, imm, measurement, queue, Supply Chain, system, waiting time. Bookmark the permalink.

19 Responses to The challenges in estimating health system response to waiting time measurement systems

Patrick Lee says:

February 16, 2016 at 3:37 pm

Yes referrals should be tracked. Deaths have occurred as a result of long wait times. A measurement to use should be the number of patients waiting for a mental health treatment. Also, incorporate a measure showing the patient was actually seen and began treatment–would capture patients’ perspective. The incentives for this would not work—data would be falsified show a very low waiting period, when most likely, it was and is higher.

Reply
Emily London says:

February 16, 2016 at 5:46 pm

*How should the system be measured to decrease the incentive to create unintended consequences?

The system should be measured from the date and time of the initial report from patient to hospital, date and time the patient has first appointment, and the date and time the patient’s issue is resolved (if resolved).

*Should the referral process be tracked across doctors and clinics to ensure end-to-end measurements?

Yes, I do not understand why doctors and clinics themselves would not measure their referral process. If they really cared to save people’s lives, they should always be striving to improve their operations.

*How should the patient’s perspective be ensured in these measurements?

A patient survey and if the doctors and clinics score low, an investigation should be made.

*Could incentives be designed that will ensure prompt service while maintaining quality?

Just a good score and reputation to identify better quality facilities. It is sad that a doctor or clinic would need incentives to properly track and serve patients, especially the VA.

Reply
Frank Griffin says:

February 17, 2016 at 6:52 am

Consistency needs to be a part of the system. Possibly a industry measurement format, to eliminate ambiguity of measurements should be implemented. Also this system should be a implemented across all doctor offices. As for any customer service operation, the patient’s perspective should be considered since the patient is the customer. Unfortunately supply and demand does not always drive the right results in the health care industry (my opinion). Supplying sub-quality service in health care causes complaints but many times do not cause a shift in doctor office demand, because of switching costs, insurance limitations, etc. So some other incentives need to be put into place to force that waiting time is measured consistently, accurately and does not affect overall health of patient.

Reply
Paul C. Barron says:

February 17, 2016 at 10:45 am

Consistency should absolutely be established for the “waiting time measurement system” (agree with Frank). Whichever method is used, it should be designed in such a manner that it can’t be manipulated by health system administrators, it should be capable of producing publicly accessible reports, and (most importantly) it should include the patients’ real experience. By “real experience” I mean the measurements should be made based on interaction points, such as:
– Initial appointment made to meeting with healthcare professional and either issue resolution, referral assignment, follow-up, etc.;
If issue has not be resolved, then:
– Referral/follow-up appointment made to meeting with that next healthcare professional and either issue resolution, additional referral, additional follow-up, surgery…whatever…etc.;
– Just keep these cycles up to final issue resolution.
(I’m not going to go into the technical means this could be done, but with technology as it is, cloud based computing, and means for encrypted individual identification–I’m sure someone could figure out a good system–maybe Google would do it pro bono!)

Whether the healthcare professional is a doctor within a larger-healthcare system, an independent nurse practitioner, or a clinic, the measurement system should track all of these interactions.

I don’t like the idea of offering incentives. After all, who will provide these incentives–the federal government? That would mean that healthcare systems would have inflows from the government, health insurance, and patients. Starbucks doesn’t receive incentives (outside of increased patronage) for decreasing wait times, so why should a hospital.

Reply
Sooin Kim says:

February 17, 2016 at 7:00 pm

Medical resources, especially scarce expert resources should be efficiently used. The question would be how to build a “sorting” system without adding an additional bureaucratic layer, and free up some core capacity for the real emergency situation. The medical system seems to be rather different from the US, Europe and Asia but a home doctor can obviously play this role in general observing the system in the Netherlands. A recent article from the Netherlands for instance said about how many “emergency” facilities are occupied by “non-emergency” cases in reality. A home-doctor post at the emergency hospital can also sort very quickly if it requires specialists treatment.
Furthermore, the insurance coverage should not be misused by increasing medical treatment unnecessarily. While it is important for an insurance policy to cover key diseases or must-have medical treatments, it is also important to develop medical products and premium policy such a way that people get incentives by not using medical services by lowering premiums, and also charge a certain minimum amount for each time medical services are used. This way, many pseudo emergency cases can be minimized, and safeguard the 24 or 48 hour emergency services. Obviously, there should be an objective or mutually agreed waiting time information between consumers and medical professionals available to both hospital and the public.

Reply
M. Moore says:

February 24, 2016 at 4:25 am

How should the system be measured to decrease the incentive to create unintended consequences?

The measure for the Key Performance indicators (KPI) should have a standard start time (initial visit), this will ensure a base starting line. In addition to measuring if a patient receives healthcare services, a bookend KPI that measures patients taking greater than 48 hours should be considered. This measurement with a mean time beyond the goal should be tracked to identify to what extent and why some patients go beyond the 48 hrs. before receiving care .

Should the referral process be tracked across doctors and clinics to ensure end-to-end measurements?
This measure would also be important to identify the lag time to get the patient treated when they are not able to be treated and referred to a specialist.

How should the patient’s perspective be ensured in these measurements?
The patient’s perspective will benefit if the visit is expedited and the treatment provided on the initial visit. The measurement should also measure if the patients issue was addressed/resolved in the minimum # of visits to ensure best quality of care.

Could incentives be designed that will ensure prompt service while maintaining quality?
Incentives could be developed to reward those institutions that meet a goal for prompt service and factor in the quality of the service component.

Reply
Matt Geddie says:

February 24, 2016 at 6:57 pm

Yes, an overall tracking system may be daunting but creating a system in which those who recieve care outside the 48hr range are marked, leading to penalties faced by the care providers.

Yes, tracking the referral process from beginning to end to ensure fairness across the board and eliminate bias.

If the patient receives treatment on their initial visit their perspective will greatly increase. Furthermore, the care provider should make it a priority to address the patient’s issue, further increasing the patient’s perspective.

Creating a system that treats patients in an efficient time manner, whilst also weighing the patient’s personal experience through a survey would be two good metrics to base an incentivization structure on.

Reply
Ryan Laskey says:

February 25, 2016 at 8:53 am

I would question the difference between of 95% and the 64%. I would imagine that the VA is measuring their fill rates by the percentage of patients who they have scheduled appointments for and the 64% are the patients who have actual had their appt with the doctor. If this were the case , then 30% of people waiting in queue for the doctors. There is little to no queue ahead of scheduling.

Reply
Mike Flatt says:

February 25, 2016 at 4:06 pm

I had not been concerned with the disparity of the data until Ryan pointed it out. It provides certain cause for concern that either the wrong data is being measured or perhaps it is being manipulated by either side of the argument to their benefit. As with any measurement system there will always be some element of manipulation and gaming of the system, however that should not stop us from attempting measurement. Personally I would measure it like inventory days on hand. The providers want to serve patients and get paid so they are incentivized on throughput. If the people in the queue are taken in relation to past throughput performance over a given time the “days on hand inventory” or similar comparison and an appropriate reduction goal should drive the desired behavior.

Reply
Dan Skinner says:

February 25, 2016 at 8:15 pm

The system should be measured from the beginning of the referral process, to the time of first appointment, to seeing a specialist (if required), to receiving treatment.
I think that disparities of VA reported times to actual times needs to be dealt with harshly to make an example. They are committing fraud against the US government.
The incentives do need to be rethought, but the current metrics only encourage people gaming the system. I do not know what might work.

Reply
Sandra Aldana says:

February 25, 2016 at 10:47 pm

I agree with Dan that the system should be measured from the customer’s perspective. However, I disagree with the measurement. I think the clock should be ticking when the person attempts to make the appointment with the specialist. Often times, the doctor makes the referral but it might take days until the person calls to make their appointment. That time should not be counted as part of the process.

Reply
LaBaron Hartfield says:

February 26, 2016 at 10:06 am

“Any waiting-time measure can be thwarted or misrepresented,” says Michael Davies, an internist and acting director of high reliability systems and consultation at the U.S. Department of Veterans Affairs.

I have to question should why would we ever want to incentivize medical care based on throughput metrics. Creating a fair and equitable measurement that captures the response times could degrade the quality care based on the rigidity that would have to be applied to processes to capture the data needed. There are several externalities that affect the measurement that physicians do not control.

The VA has a better chance to create such as structure as it is an integrated closed loop system – the VA supplies or contracts the majority of it services. Within a closed loop is an opportunity to link an initial activity with a follow-up activity. In the private healthcare market, physicians are not always able to refer patients to specialists within their network. Referring physicians also must contend with the constraints generated by their patient’s private insurance plans. This “system wait” must be captured and extracted from the measurements.

Additionally, the metrics must capture and omit the high levels of variability that exist due to patient action or inaction. The patient must schedule and attend an appointment based on the referral to complete the cycle. Even once scheduled, appointments can rescheduled by the patient (or provider) which effectively resets or extends the clock on the measurement of the “system wait.” The patient receives no punishment for this flexibility so has not incentive to keep the appointment, however certain interpretations of the time could punish providers who will react by gaming the system to work more favorably towards their rewards.

Reply
Courtney Metzger says:

February 28, 2016 at 6:22 pm

Quick story – I was involved in a very serious accident and I had to wait nearly a month to see a Neurosurgeon. The first Dr. I met with wanted to send me into surgery the next day. I went for a second opinion (an additional 3 week lead time) and another Neurosurgeon encouraged me to continue physical therapy. Long story short, although I have some remaining symptoms, ultimately surgery would have been a poor choice for me.

The reality is that many specialist make their money on the operating table. If the incentive for face-to-face time was more competitive with the pay-offs offered in the operating room, perhaps throughput would be increased and a reduction in unnecessary surgeries would be the result.

Reply
Selva Vadivelmurugan says:

February 29, 2016 at 10:51 am

Waiting time should be measured by how critical (i.e. life threatening or irrecoverable) the issue was or for any procedure that is deemed as not cosmetic

It is good to keep track of referral process because it might help save from repeating procedures if it was already carried out by a certified professional elsewhere. Some hospitals seem to repeat such procedures as part of their protocol which I think is not only waste of money but also time.

Patients should be informed of such delays with reasons on why they have to wait so much more time. If there is possibility to expedite, then they should be given such an option where they can pay more money to get a priority service (provided it does not impact others with critical/irrecoverable illness). Such options could help hospitals to recruit more staff to provide such service thus maintaining quality as well.

Reply
Oswin Joseph says:

February 29, 2016 at 4:34 pm

Healthcare system needs to move from paying for number of services – many of which are not necessary to the speed and quality of only needed services. An end to end measurement of all referrals can be useful in reducing the number of unnecessary procedures and unintended consequences as doctors and hospitals will be less inclined to refer patients needlessly if they think they are being tracked and someone may ask questions. The VA seems to have historical trouble tracking waiting times so there seems to be a fundamental intentional fudging of the numbers than a difference of opinion from the customers.

Reply
Meera Gursahaney says:

March 3, 2016 at 3:34 pm

The system should be measured from the first appointment to diagnosis and from diagnosis to beginning the remediation (surgery, medicine, nothing etc). Any system that involves incentives should expect the people involved to manipulate the data within the law as much as possible to meet those incentives. To combat this, strict procedures should be in place to outline the timing expectations of these metric measurements to ensure true measurements. For example, the timing of when a patient is to be added to a specialist queue should be defined. Periodic audits should take place to ensure compliance and harsh penalties should be enforced if purposeful non-compliance or data manipulation is found. The only way to ensure end-to-end measurements to to track from initial visit to remediation plan in place across doctors/clinics.

Reply
Rodney Williams says:

March 4, 2016 at 6:18 pm

I do not believe that by not having a standardized method of measuring wait times amongst health clinics, is the only reason for the excessive wait times. In my opinion the best way to reduce the wait time would be to increase capacity at the bottleneck, which would require increasing the number of doctors and specialist. This would allow the system to work more efficiently by having patients being able to be seen sooner, thus reducing the wait time. I do agree that wait time should be measured between doctors and healthcare clinics in order to capture that entire process and to also standardize the measurement across the industry. In choosing the appropriate incentive it is important to focus on one that rewards for total process improvement and not one that strictly focuses on wait time reductions. This will prevent healthcare facilities from focusing on newer patients at the expense of patients that been in queue longer due to them already exceeding wait time goals. If the overall process is the focus, then healthcare facilities will not prioritize newer patients like the current system is driving them to.

Unfortunately, I do no see a way of improving the wait times without increasing the number of specialist and doctors without bugging them down with a high quantity of new patients added to the system. This will efficiently increase the throughput of the system by getting patients through faster and reducing wait time without sacrificing quality of care.

Reply
Marcellus Robinson says:

March 4, 2016 at 6:57 pm

Every system has KPI’s. To me, it only makes sense that the most important KPI to measure is the safety of people. It is irresponsible and I would go so far as to say wreck-less for the healthcare community to not have data to understand the impact of wait times and the referral system as a whole. How can you possibly make adjustments and ensure “quality of life” services if you don’t know where inefficiencies lie.

I’m going to make a sweeping generalization that Doctors treat what is presented to them without regard for who is waiting to be treated. I give them an “out” because they can’t have to time to be aware of both. But their profession is unique in the context that unlike other professions, days, minutes and seconds absolutely matter when you are talking about potential fatal or long-term implications of not receiving timely treatment. It is absolutely important that an adoptable system is created so we, as a society, can begin to understand where we are going by understanding where we have come from. How, by possessing actionable and applicable data.

Reply
C. Thomas says:

March 7, 2016 at 7:41 pm

How should the system be measured to decrease the incentive to create unintended consequences?

We should be careful how we incentivize doctors regarding wait times. While it is important it should be given the appropriate weight based upon other important factors as well such as patient outcomes. If we reward the healthcare system for only seeing patients quickly the system will find a way to offer quick follow-up appointments while the quality of that appointment is likely to decrease.

Should the referral process be tracked across doctors and clinics to ensure end-to-end measurements?

For a facility like the VA in which there are a vast array of physician resources for any given condition I believe a pooling process would greatly enhance the efficiency of the system to reduce the lead time of the queue due to the fact that available doctors would never be idle if there is a queue. While patients may not always see the same doctor, enhanced communication from doctor to doctor via the prevalence of electronic medical records has been shown to make up for that. There are many patients with conditions that warrant faster turnaround time at the expense of seeing a familiar physician.

How should the patient’s perspective be ensured in these measurements?

The patient perspective should be ensured by including patient satisfaction scores and rewards based on performance in terms of patient outcomes.

Could incentives be designed that will ensure prompt service while maintaining quality?

Yes, if the system is appropriately weighted to force physicians to focus on multiple quality avenues as previously described.

Reply