ORIGINAL ARTICLE HEALTH SERVICES RESEARCH AND POLICY

Shorter Perceived Outpatient MRI WaitTimes Associated With Higher PatientSatisfaction

Anna Holbrook, MDa, Harold Glenn Jr, BS a, Rabia Mahmood, MPHa, Qingpo Cai, BSa,Jian Kang, PhDb, Richard Duszak Jr, MDa

Abstract

Purpose: The aim of this study was to assess differences in perceived versus actual wait times among patients undergoing outpatientMRI examinations and to correlate those times with patient satisfaction.

Methods: Over 15 weeks, 190 patients presenting for outpatient MR in a radiology department in which “patient experience” is one ofthe stated strategic priorities were asked to (1) estimate their wait times for various stages in the imaging process and (2) state theirsatisfaction with their imaging experience. Perceived times were compared with actual electronic time stamps. Perceived and actual timeswere compared and correlated with standardized satisfaction scores using Kendall s correlation.

Results: The mean actual wait time between patient arrival and examination start was 53.4 � 33.8 min, whereas patients perceived amean wait time of 27.8 � 23.1 min, a statistically significant underestimation of 25.6 min (P < .001). Both shorter actual and perceivedwait times at all points during patient encounters were correlated with higher satisfaction scores (P < .001).

Conclusions: Patients undergoing outpatient MR examinations in an environment designed to optimize patient experience under-estimated wait times at all points during their encounters. Shorter perceived and actual wait times were both correlated with highersatisfaction scores. As satisfaction surveys play a larger role in an environment of metric transparency and value-based payments, betterunderstanding of such factors will be increasingly important.

Key Words: Wait times, patient satisfaction, value-based payments, quality metrics, MRI

J Am Coll Radiol 2016;13:505-509. Copyright � 2016 American College of Radiology

INTRODUCTIONPursuing a variety of quality- and value-based paymentinitiatives, CMS began a major shift in its approach toprovider reimbursement. Two years ago, it announcedplans to transition its traditional fee-for-service hospitalpayment system to one that bases reimbursement onquality and value [1]. More recent announcements havetargeted similar efforts for physician reimbursement [2].

This new focus on value-based purchasing increas-ingly directs attention to the patient experience. Hospital

aEmory University, Atlanta, Georgia.bUniversity of Michigan, Ann Arbor, Michigan.

Corresponding author and reprints: Anna Holbrook, MD, Emory Uni-versity School of Medicine, 1365C Clifton Road, NE, Building C, SuiteC1104, Atlanta, GA 30322; e-mail: anna.holbrook@emoryhealthcare.org.

This study was funded by the Beryl Institute Patient Experience GrantProgram. The authors have no conflicts of interest related to the materialdiscussed in this article.

ª 2016 American College of Radiology1546-1440/15/$36.00 n http://dx.doi.org/10.1016/j.jacr.2015.11.008

Consumer Assessment of Healthcare Providers and Sys-tems surveys assess patients’ perspectives of care [3].Although their association with clinical outcomes hasbeen challenged [4], such surveys are now beingused by CMS as a component of hospital incentiveprograms [5].

Although wait times are not specific components ofHospital Consumer Assessment of Healthcare Providersand Systems surveys, they could influence patients’ re-sponses to a variety of generic survey questions [3]. Oftenattributed to an undesired but normal and sometimesunavoidable aspect of hospital operation, wait timesplay an influential role in the overall patient experienceand begin far before patients interact with their clinicalproviders. Longer wait times contribute to patientanxiety and dissatisfaction and are frequently citedas a common reason for patients’ leaving a medicalpractice [6].

505

Findings in Press Ganey’s [6] 2009 Medical PracticePulse Report indicate, not surprisingly, that overallpatient satisfaction declines as wait times increase. Yetpatients typically respond to such surveys on the basisof their perceptions of wait times, rather than actualmeasured wait times. In nonehealth care markets,consumer estimates of wait times have been shown tobe notoriously unreliable [7-9]. A better understandingof differences between the subjective and objectivecould help providers and hospitals improve patientsatisfaction and ultimately their bottom lines. TheACR’s recent launch of the new Commission onPatient Experience [10] reflects a heightened awarenessby the specialty of issues related to satisfaction, as wellas opportunities for targeted research in this arena.

For these reasons, we aimed to assess differences inperceived and actual wait times among patients and tocorrelate these times with standard satisfaction scores,focusing on outpatient MRI services at our tertiary careacademic medical center.

How would you rate your experience in Radiology today?1 2 3 4 5

Very Poor Poor Fair Good Very Good

How many minutes did you wait between checking in and prep?__________________________________________________

How many minutes did you wait between prep and starting the actual test?__________________________________________________

Roughly how many minutes was your actual test?__________________________________________________

How long did you anticipate being here today?__________________________________________________

While waiting, were you offered anything to pass the time?__________________________________________________

Fig 1. Patient survey.

METHODSThis HIPAA-compliant study, which was initiated as apatient experience quality improvement project, wascompleted under a waiver from our institutional reviewboard. Optimization of the patient experience is a statedstrategic goal of our academic radiology department. Forexample, all new faculty and staff members attend day-long Radiology Service Excellence Institute retreats, anddesignated patient and family advisers participate in moststrategic and capital planning sessions. As such, initiativessuch as this support our departmental goals.

Over a 15-week period (October 14, 2014, toFebruary 27, 2015) patients arriving for scheduled MRexaminations during periods staffed by our patient serviceresearch team were asked to consider participating in thisvoluntary study and given assurance that their responseswould not be shared with those involved in their care. Forthose agreeing, patient demographic information, variousencounter component times, and the type of MR ex-amination (brain, chest, abdomen, pelvis, spine, orbreast) were captured and stored in a secure serverdatabase.

Our academic health system performs approximately70,000 MR examinations per year at its 10 hospitals andoutpatient imaging centers. Of these, 17,000 are per-formed on inpatients, 51,000 on outpatients, and 2,000on emergency department patients. For this study, wefocused on the outpatient MR facility at our primary

506

university campus, which has two MR units and a vol-ume of approximately 11,500 examinations per year. Ouroutpatient MR waiting room was specifically designedwith the patient experience in mind, with ample readingmaterials, multiple large-screen televisions, and freeWi-Fi. Team members checked on patients periodicallythroughout their stays, offering progress updates, warmblankets, and coffee or water. Additionally, the researchteam was instructed to offer half of the patients iPadelectronic tablet devices preloaded with games andInternet access as a wait distractor.

Upon the completion of their MR examinations,patients were asked to complete a five-question survey(Fig. 1), which, on the basis of initial pilot testing, usuallytook between 1 and 3 min. The survey was administeredon paper. In five cases, however, because of poor vision orneed for an interpreter, questions were asked by a studyteam member and answered verbally. Patients wereasked to rate their experience in the outpatientradiology department on a five-point, Likert-type scale,modeled after those in a typical Press Ganey patientsurvey. We asked how long they thought they waitedbetween their arrival and when they were brought back tothe MRI preparation area and how long they waited afterbeing brought to the preparation area until their exami-nations began. These two perceived wait period compo-nents were combined to calculate the patient’s perceivedtotal wait time. We also asked patients to estimate howlong their examinations lasted, how long they hadanticipated being in the department, and if they had beenoffered anything by the research or clinical team to passtheir time.

Actual appointment time, time of arrival, and time ofexamination start and finish were all extracted from ourinstitutional data warehouse using specific fields imported

Journal of the American College of RadiologyVolume 13 n Number 5 n May 2016

Fig 2. Actual versus perceived wait times in the radiologydepartment, including from when the patient arrived to thetime the MRI examination started, the duration of the MRIexamination, and total time in department.

from scheduling, registration, and scanner software datasets. Actual and perceived wait times were comparedusing two-sample t tests. Actual and perceived examina-tions durations were also compared using two-sample ttests.

We also compared how actual and anticipated totaltime spent in the department and the difference betweenthe two correlated with reported satisfaction using Ken-dall s correlation tests. We compared how the type ofexamination correlated with actual and perceived waittimes or patient satisfaction using Kendall s correlationtests. Finally, we compared how the offering of an iPad asa distractor affected perceived wait times, satisfaction, andthe difference between perceived and actual wait timesusing two-sample t tests. All data and statistical analysiswas performed using Excel 2010 (Microsoft Corporation,Redmond, Washington) and R (The R Foundation forStatistical Computing, Vienna, Austria).

Table 1. Frequency of satisfaction scores from patient survey

Score 1 2 3 4 4.5 5Frequency 2 0 1 24 4 116

RESULTSOf 190 patients invited to participate, 11 declined, and32 left the department before completing the surveyportion of the study, yielding a sample of 147 patients forwhom complete data were available. Of these, 73 patients(49.7%) were offered something by our research team topass the time.

The mean time intervals between patient arrival andexamination start, time of appointment and examinationstart, and time of arrival and time of appointment were53.4 � 33.8, 20.2 � 53.6, and 33.0 � 45.4 min,respectively. Patients perceived a mean of 12.7 � 10.7min elapsing between their time of arrival and theirpreparation and a mean of 15.7 � 18.6 min elapsingbetween their preparation and the examination begin-ning, for a mean total perceived wait time of 27.8 � 23.1min. This represents an underestimation compared withactual wait time. The difference between perceived totalwait time and the actual interval between arrival time andtest start was statistically significant (P < .001), but therewas no significant difference between the patients’perceived wait times and the interval from time ofappointment to time of test start (P ¼ .116). Differencesbetween actual and perceived component and total waittimes are illustrated in Figure 2.

The median patient satisfaction score was 5 (Table 1).There was a statistically significant correlation betweenpatient satisfaction scores and each component of actualand perceived wait times, with shorter wait timesassociated with higher satisfaction scores (actual arrival

Journal of the American College of RadiologyHealth Services Research and Policy n Holbrook et al n MRI Wait

to test start, P < .001; appointment to test start, P ¼.002; perceived wait time between arrival andpreparation, P < .001; perceived wait time betweenpreparation and examination start, P < .001; andpatient total perceived wait time, P < .001).

The mean examination duration was 47.1 � 22.7min, but patients estimated the duration of their exami-nations to be 31.6� 15.2 min. This underestimation wasstatistically significant (P < .001).

Mean anticipated time in the department was 71.7 �41.2 min. Mean actual time was 100.3 � 42.3 min. Thedifference between the anticipated and actual times(27.5 � 49.1 min) was statistically significant (P < .001)and was correlated significantly with patient satisfactionscores (P ¼ .005). Mean perceived time in the depart-ment was 59.7 � 29.6 min. The difference between theanticipated and perceived times (13.8 � 44.6 min) wasalso statistically significant (P < .001). Smaller differencesin anticipated and perceived time were associated withhigher patient satisfaction scores (P ¼ .007).

The difference between actual (100.3 � 42.3 min)and perceived (59.7 � 29.6 min) times was 40.2 � 36.9min, which was also statistically significant (P < .001).Differences between actual and perceived times in thedepartment were not significantly correlated with satis-faction (P ¼ .06).

There was no significant association between the typeof examination performed and the actual time of arrival to

507Times and Patient Satisfaction

examination start (P ¼ .436), time of appointmentto examination start (P ¼ .060), perceived wait times(P ¼ .796), or patient satisfaction score (P ¼ .376).

If a patient was offered an iPad tablet device to passthe time, it did not affect the patient’s perceived wait time(P ¼ .408) or satisfaction score (P ¼ .791) or the dif-ference between his or her perceived wait time and theactual wait time (P ¼ .588). There was no significantassociation between a patient’s being early or late for hisor her examination and the patient’s perceived wait time(P ¼ .006).

DISCUSSIONContemporaneously capturing perceived and actual waittimes for patients undergoing outpatient MR in anenvironment designed to optimize the patient experience,we found that patients underestimated their totalencounter times by almost 30 min. For all components ofthe imaging encounter, shorter wait times, both perceivedand actual, were significantly associated with higherencounter satisfaction scores. As the patient experiencebecomes an increasingly important component of publicratings websites and value-based reimbursement systems,this information may help radiology practices better aligntheir workflows with societal imperatives.

Prior research in the nonradiology setting has shownthat when patients received clinical information duringtheir waits, they perceived shorter lengths of stay [11].Because providing real-time clinical information to pa-tients who are presenting for and undergoing diagnostictesting is impractical, our department has focused onproviding whatever information possible (eg, informationabout the examination, anticipated delays) in ascomfortable and patient-centric an environment aspossible during each imaging encounter. Given theobservational nature of our study (eg, we did notrandomize patients to “old-school” institutional waitingenvironments), we cannot assert causation that thesefactors led patients to think that wait times were shorterthan they really were. But our positive results in a patient-centered waiting environment may explain our favorableexperience.

Prior studies comparing perceived and actual waittimes in the health care setting are few and have yieldedmixed results. Not unexpectedly, some researchers havefound that patients overestimate wait times. Thompsonet al [12], for example, found that almost half ofemergency department patients overestimated the timefrom triage to examination by a physician. Others,

508

however, have reported, like us, that patientsunderestimated their wait times. In particular, Parkerand Marco [13] reported that the majority ofemergency department patients underestimated theirvisit lengths by a median of 7 min. The explanation forthis discordance could be a simple one. Rather thanone report being right and the other being wrong, withpatients always either over or underestimating waittimes, we believe it is likely that patients estimate oneway or the other on the basis of the environment inwhich their care was delivered.

In an era in which patient satisfaction scores areincreasingly serving as value-based reimbursement metrics[3], efforts to improve satisfaction will increasingly affectphysician and facility bottom lines. Prior studies havereported associations between shorter perceived andactual wait times and improved satisfaction scores inthe emergency department [13] and ophthalmologyclinic [14] settings. Our study indicates that thesereports can be generalized to the outpatient MRIsetting as well. Practices may be able to improve actualwait times only so much, so focusing in parallel on thepatient experience—and reducing the perceived lengthsof those times, whatever they are—may yieldcomplementary end results. In fact, perception may bemore important here than reality; in one clinic waitingroom setting, patient reactions were actually morestrongly affected by perceived rather than actual waittimes [15].

We had expected that offering an iPad tablet devicefor patients to pass the time might shorten their perceivedwait times and improve their satisfaction. Like Pruyn andSmidts [15], who did not find the presence of a televisionin the waiting room to decrease patients’ perceived waittimes, we did not find this to be the case. The reasonfor this is unclear but could be related to manypatients’ already having their own devices (somethingwe had not expected and thus did not ask about). Itcould also be due in part to the fact that ourdepartment had already succeeded, through both staffdevelopment and facility design, to make the waitingexperience as pleasant as possible. Because of both, theincremental value of distractors may have beenmitigated. For these reasons, it is also not clear whetherour findings that patient underestimate wait times willnecessarily generalize to all radiology departments. Fordepartments that find patients overestimating their waittimes, however, some of our patient experienceinitiatives might be considered if corresponding patientsatisfaction metrics are suboptimal.

Journal of the American College of RadiologyVolume 13 n Number 5 n May 2016

There were several limitations in our study. First,because this was a single-institution study focusing on asingle site of service, these results may not necessarily begeneralizable to other environments (eg, freestandingoutpatient imaging centers), which are structurally andorganizationally very different from our academicmedical center practice. Second, our team’s other obli-gations precluded them from being able to be presentduring the early morning (6-9 AM). Approximately 33%of patients were scheduled during these hours and weretherefore not represented in this study, but we have noreason to believe that this group would be different intheir opinions from other survey respondents. Addi-tionally, several patients who had been recruited for thestudy did not receive the survey. This was due to factorssuch as patient fatigue, tardiness to other same-dayappointments, and patients’ leaving forgetting thatthey had agreed to participate. We do not know whatbias this may have introduced. Finally, patient partici-pation was somewhat inversely proportional to sched-uling capacity. When scanners were running onschedule, patients were taken back quickly for theirexaminations, and such patients were thus less likely tobe enrolled. Conversely, when processes were runningbehind schedule (eg, because of a large number ofemergency add-ons to the schedule), more patients weresurveyed. This biased our study toward including pa-tients with longer waits, suggesting we may have seenmore favorable results if we were able to capture a morebalanced sample.

JH

TAKE-HOME POINTS

- In a modern patient-centric facility, patients un-dergoing outpatient MR tended to underestimatewait times.

- Shorter actual and perceived wait times were bothcorrelated with higher patient satisfaction scores.

ournal of the American College of Radiologyealth Services Research and Policy n Holbrook et al n MRI Wait

- As satisfaction surveys play a larger role in emerginghealth care delivery systems, an increased emphasison the patient experience will likely augment cur-rent objective time-based metric assessments.

REFERENCES1. Centers for Medicare and Medicaid Services. Linking quality to payment.

Available at: http://www.medicare.gov/hospitalcompare/linking-quality-to-payment.html?AspxAutoDetectCookieSupport¼1. Accessed July 1, 2015.

2. Burwell SM. Setting value-based payment goals—HHS efforts toimprove U.S. health care. N Engl J Med 2015;372:897-9.

3. Centers for Medicare and Medicaid Services. Hospital ConsumerAssessment of Healthcare Providers and Systems. Available at: http://www.hcahpsonline.org. Accessed July 1, 2015.

4. KennedyGD,Tevis SE,KentKC. Is there a relationship between patientsatisfaction and favorable outcomes? Ann Surg 2014;260:592-8.

5. Centers for Medicare and Medicaid Services. HCAHPS: patients’perspectives of care survey. Available at: http://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/HospitalQualityInits/HospitalHCAHPS.html. Accessed July 1, 2015.

6. Press Ganey. Keep me waiting: medical practice wait times and patientsatisfaction. Available at: http://www.pressganey.com/Documents_secure/Medical Practices/White Papers/Keep_Me_Waiting.pdf. AccessedJuly 1, 2015.

7. Maister DH. The psychology of waiting lines. In: Czepiel JA,Solomon M, Surprenant CS, eds. The service encounter. Lexington,Massachusetts: Lexington Books; 1985:113-23.

8. Jones P, Peppiatt E. Managing perceptions of waiting times in servicequeues. Int J Serv Ind Manag 1996;7:47-63.

9. Durrande-Moreau A. Waiting for service: ten years of empiricalresearch. Int J Serv Ind Manag 1999;10:171-89.

10. Allen B. Introducing the ACR Commission on Patient Experience2015. ACR Bulletin. Available at: https://acrbulletin.org/august-2015/254-boc-patient-experience-acr. Accessed September 9, 2015.

11. Tran TP, Schutte WP, Muelleman RL, Wadman MC. Provision ofclinically based information improves patients’ perceived length of stayand satisfaction with EP. Am J Emerg Med 2002;20:506-9.

12. Thompson DA, Yarnold PR, Adams SL, Spacone AB. How accurateare waiting time perceptions of patients in the emergency department?Ann Emerg Med 1996;28:652-6.

13. Parker BT, Marco C. Emergency department length of stay: accuracyof patient estimates. Western J Emerg Med 2014;15:170-5.

14. McMullen M, Netland PA. Wait time as a driver of overall patientsatisfaction in an ophthalmology clinic. Clin Ophthalmol 2013;7:1655-60.

15. Pruyn A, Smidts A. Effects of waiting on the satisfaction with the service:beyond objective time measures. Int J Res Mark 1998;15:321-34.

509Times and Patient Satisfaction