Factors Associated with No-Show Rates in a Pediatric Audiology Clinic

Abstract

Objective:

To evaluate factors associated with no-show rates in a pediatric audiology clinic.

Study Design:

Retrospective review.

Setting:

Tertiary referral center.

Participants:

All pediatric patients under the age of 18 whose parents/guardians scheduled an appointment at a tertiary Audiology Clinic between June 1, 2015 and July 1, 2017.

Main Outcome Measure(s):

Data included whether the patient came to their appointment, patient age, sex, race, insurance type, appointment type, location, season of appointment, and day of the week of the appointment.

Results:

Of the 7,784 pediatric appointments scheduled with audiology, the overall no-show rate was 24.3% (n=1893). Lower age was significantly associated with no-shows (p=0.0003). Black/African American children were more likely to no-show compared to White/Caucasians (p=0.0001). Compared to self-pay/military/other insurance, those with Medicaid were more likely to no-show (p=0.0001). The highest rate of no-shows occurred during summer (27%). On multivariate analysis, younger age, Black/African American race, and Medicaid insurance were associated with increased no-show rates.

Conclusion:

A variety of factors influence no-show rates in a pediatric audiology setting. No-shows can impact treatment quality and affect overall hearing outcomes. Further investigation is necessary to assess barriers to appointment adherence and to develop interventions to improve adherence and care.

Introduction:

The designator “no-show” is an indicator of a health behavior defined as not attending a scheduled appointment without making a cancellation.¹ Nonadherence to medical appointments can impact patient care in a variety of ways. Previously, it was demonstrated that patients with a higher propensity to miss appointments have had increased rates of acute care utilization, hospitalization, and emergency department visits.²

The prevalence of no-show behavior is well-documented across multiple medical settings. Kheirkhah et al. studied no-shows in a large VA Medical Center with retrospective data over a 12-year period and reported an overall no-show rate of 18.8%.³ Specifically, the VA audiology clinic reported a 12.6% no-show rate, resulting in an estimated of $10.48 million per year loss to the medical center.³ Fiorillo et al. studied no-show rates in a large otolaryngology practice over a 1-year period and reported an overall rate of 20%.⁴

Not only do no-shows result in a reduction of revenue for medical centers, but also represent a missed opportunity for patient care. In the case of pediatric audiology, prompt diagnosis and treatment of hearing loss are essential for children’s language, social, and cognitive development.² Therefore, identification of factors that may predict nonadherence to scheduled appointments in this particular patient population may improve outcomes if addressed early with effective strategies such as patient navigator programs⁵ or electronic reminders.^6–8

In the literature, distinct factors are proposed as predictors for no-show behavior. Demographics such as male sex, ^9–13 younger age, ^9,14–19 minority ethnicity, ^10,12,13,19 and lower socioeconomic status, ^{12,13,18,20,21} have shown to increase no-show rates in adult sample populations. Additionally, a lack of transportation, ^22–25 long travel distances to the medical center, ^21,26 and government-issued or lack of health insurance, ^{11,12,14,17,19–21,27} have also been shown to have an association with higher no-show rates. No-show rates tend to rise during the winter months ^20,27,28 and even during particular days of the week ³ or at certain times of the day.^9,11 Marital status, lead-time to appointment, time since referral, and previous rescheduling of appointments were also shown to affect no-show rates.^{9,14,16,19,28,29} Most of the literature in this area has evaluated adult no-show rates. Fiorillo et al. demonstrated that there was no difference in the overall no-show rates between the adult and pediatric patients in the academic otolaryngology setting. This study did not address whether the factors influencing no-show rates in adults were consistent with the predictive factors in pediatric patient populations.⁴

Despite the considerable work done in other fields, factors associated with increased rates of appointment nonadherence in the pediatric audiology setting have not yet been robustly described. The aim of this study is to identify factors associated with no-shows in a pediatric audiology clinic. Describing a population vulnerable to high no-show rates opens the field for future research on distinct interventions to address missed appointments and mitigate the adverse effects of delayed diagnosis and intervention in pediatric audiology populations.

Material and Methods:

Design and Population

This is a retrospective study involving pediatric patients whose parents/guardians scheduled an appointment to see a provider at a university-based tertiary care audiology practice between June 1, 2015, and July 1, 2017. The majority of patients were seen at an suburban, hospital-based clinic located within the southeastern Midwest region. A smaller proportion of patients were seen at university sponsored satellite clinic located within a one-hour drive from the primary clinic. The appointment information was obtained from two sources: the outpatient electronic medical record system (Allscripts Electronic Health Record -AEHR-) and the billing system (Allscripts Practice Management -APM-). Missed data was obtained through a chart review using the AEHR to assure completeness. Sample size determination is illustrated in Figure 1. This study was approved by the Institutional Review Board of the Office of Research Integrity of the primary institution (number 15–0861-P3H).

Figure 1.

Sample Size Flowchart of Pediatric Audiology Appointments between June 2015 and July 2017.

No-show appointment (outcome of interest)

Appointments were classified as a dichotomous variable: show or no-show. A no-show was defined as an appointment in which individuals neither attended nor canceled/rescheduled the scheduled appointment.

Risk factors (predictor variables)

Based on the literature review, the following information was included as a potential predictor for a no-show appointment: age (continuous variable measuring age in years), sex (dichotomous variable for male or female), race (categorical variable with three levels: White/Caucasian, Black/African American, and Other), insurance type (categorical variable with 3 levels: commercial, Medicaid, and self-pay/military), appointment type (categorical variable with 6 levels: hearing test, ABR, cochlear implants, hearing aids, auditory processing, and videonystagmography), location (dichotomous variable: hospital-based and satellite), season of the appointment(categorical variable with 4 levels: winter, spring, summer, and fall), and day of the week of the appointment (categorical variable with 5 levels: Monday-Friday).

Statistical Analysis

Explanatory analysis included the use of descriptive statistics. Specifically, frequencies and percentages are presented for categorical variables and means and standard deviations for continuous variables. Group comparison was performed using Chi-square test for categorical variables and independent t-test for continuous variables. Subjects could have as many observations as they had appointments during the study period. As such, all the analyses accounted for the correlation within subjects. To assess which predictors were statistically associated with the outcome of interest (no-show), bivariate and multivariate cluster-weighted marginal generalized estimating equation models with robust standard errors were performed. Those predictors that reached a 0.10 significance level were included in the adjusted multivariable marginal model. To fit a parsimonious model, we used a backward elimination method with a significance level of 0.05. All tests were two-sided at the 0.05 significance level. All analyses were conducted in SAS version 9.4 (SAS Institute, Carry, NC).

Results:

There were 7,784 pediatric appointments scheduled between June 2015 and July 2017 (2015 [n=1743], 2016 [n=4095], 2017 [n=1946], see Figure 1). The overall combined no-show rate was 24.3% (n=1893). Satellite clinics accounted for 7.7% (n=597) of the appointments and had a slightly higher but not statistically significant no-show rate when compared to the hospital-based clinic (25.8% vs. 24.2%, χ²=0.77, p=0.38).

Table 1 presents the descriptive statistics for the individuals and the appointments separated by the outcome of interest. Lower age was significantly associated with an increased likelihood of appointment no-show (p=0.0003). Regarding race, Black/African American children were more likely to no-show compared to White/Caucasians (p=0.0001). Regarding insurance type, the majority of patients included in the sample had Medicaid insurance. Compared to self-pay/military/other insurance, those with Medicaid were more likely to no-show (p=0.0001). In terms of appointment characteristics, most appointments were scheduled for hearing tests (n=6343 (81.5%)). Those scheduled for auditory processing and cochlear implant evaluations were less likely to no-show compared to those scheduled for hearing tests (p=0.0001, p=0.015, respectively). The highest rate of no-shows occurred during the summer (27%). Compared to the winter, patients with summer appointments had a higher likelihood of no-showing (p=0.0009) Again using the winter as a reference, patients with spring appointments had a lower likelihood of no-showing (p=0.03). The highest rate of no-shows during the week was on Mondays (24%). However, no statistically significant differences were noted between the weekdays.

Table 1.

Individual and audiology appointment characteristics. June 1, 2015 to July 01, 2017.

Patient Characteristics
	No-Show N=1893	Show N=5891	P-value
	Mean (%)	Mean (%)
Age (in years)	4.80 (4.34)	5.17 (4.47)	0.0003*
	N (%)	N (%)
Sex
Male	1133 (59.9)	3386 (57.5)	Ref
Female	760 (40.1)	2505 (42.5)	0.067
Race
White/Caucasian	1437 (78.4)	4973 (84.4)	Ref
Black/African American	324 (17.7)	665 (11.3)	0.0001*
Other	28 (1.5)	132 (2.2)	0.592
Insurance type
Medicaid	1588 (84.9)	3704 (62.9)	0.0001*
Commercial	273 (14.6)	1972 (33.5)	0.049
Self-pay/Military	9 (0.5)	215 (3.6)	Ref
Characteristics of the audiology appointments
	No-show N=1893	Show N=5891
	N (%)	N (%)
Appointment type
Hearing test	1589 (83.9)	4754 (80.7)	Ref
Auditory Brainstem Response	248 (13.1)	757 (12.9)	0.759
Cochlear implants	12 (0.6)	83 (1.4)	0.0146*
Hearing aids	30 (1.6)	202 (3.4)	0.0851
Auditory processing	7 (0.4)	79 (1.34)	0.0001*
Videonystagmography	7 (0.4)	16 (0.3)	0.588
Location
Hospital-based	1739 (91.9)	5448 (92.5)	0.457
Satellite	154 (8.1)	443 (7.5)	Ref
Season
Winter	447 (23.6)	1445 (24.5)	Ref
Spring	447 (23.6)	1705 (28.9)	0.0338*
Summer	510 (26.9)	1306 (22.2)	0.0009*
Fall	489 (25.8)	1435 (24.4)	0.1606
Weekday			0.237
Monday	454 (24)	1379 (23.4)	Ref
Tuesday	411 (21.7)	1209 (20.5)	0.992
Wednesday	396 (20.9)	1237 (21)	0.844
Thursday	395 (20.9)	1208 (20.5)	0.888
Friday	237 (12.5)	858 (14.6)	0.340

Frequencies and (percentages) are presented for categorical variables and mean (standard deviation) for continuous variables.

Table 2 presents the results for the adjusted multivariate cluster-weighted generalized estimating equation model. The initial model included 6 variables (age, sex, race, insurance, appointment type, and season). However, appointment type and sex were not significantly associated with no-shows (p>0.05) and were removed from the model. In the final model, all variables were significantly associated with no-shows. Specifically, for each one-year increase in age, the likelihood of no-show decreased (OR=0.98, p= 0.003). Black/African American children had 40% increased odds of no-showing compared to White/Caucasian patients (OR=1.41, p=0.001). Children with Medicaid had 5.68 increase in odds of no-showing compared to self-pay/military children (p<0.001). Finally, patients with appointments scheduled for the spring were less likely to have no-show appointments compared to those scheduled in the winter (OR=0.79, p= 0.026).

Table 2.

Results from the parsimonious cluster-weighted generalized estimating equation model predicting no-shows.

Variables	OR	95% CI	P-value
Age	0.975	(0.959, 0.991)	0.003
Race
Black/African American	1.411	(1.154, 1.726)	0.001
Other	1.087	(0.624, 1.895)	0.768
White/Caucasian	Reference
Insurance
Commercial	1.890	(0.838, 4.263)	0.125
Medicaid	5.675	(2.555, 12.604)	<0.001
Self-pay/Military	Reference
Season
Fall	1.070	(0.866, 1.322)	0.530
Spring	0.788	(0.638, 0.971)	0.026
Summer	1.230	(0.998, 1.517)	0.052
Winter	Reference

^*OR, odds ratio, CI, Confidence Interval.

Discussion:

This study aims to identify factors that predict no-show in a pediatric audiology clinic. Our results indicate that pediatric age, race, insurance status, and season of the appointment are significant predictors for no-show in this patient population. These findings demonstrate an opportunity to implement strategies for targeting these specific variables in order to improve clinic efficiency and patient outcomes.

Different interventions have been shown to reduce no-show rates across clinical settings. Appointment reminders via letters, text messages, and phone calls have all proven to be successful.^10–12,30 Crutchfield et al. postulate that appointment reminders have a wide range of efficacy largely due to variations in patient preferences. All patients are unique, some prefer reminders via email, text message, phone call, or a mailed letter etc. There is also considerable variability in terms of when patients would like to receive reminders.²³ In the clinics where our study was conducted, patients (or their parents/guardians) reported receiving appointment reminders via post-mail, electronic calendar reminder, email, phone call, text message, or social media reminder. Based on the published no-show rates listed previously ranging from 12 to 20%, the 24% no-show rate observed in our study sample may suggest, that these methods are not as effective in pediatric settings.

Other strategies that aim to reduce the financial burden associated with no-shows include double-booking appointment time slots. Double-booking patient is most effective when one can effectively predict the appointments at highest risk for no-show. These approaches usually utilize patient sociodemographic factors, appointment types, and appointment timing characteristics to develop strategies and identify the percent by which each day may be overbooked.^31,32 A model that utilizes the demographic identified in this study, specifically pediatric age, race, and insurance status, as a method to double-book pediatric audiology appointments should be explored in future studies. More so, minimizing the lead-time between scheduling the appointment and the actual visit date by providing same-day appointments also has delivered promising results in reducing no-show rates.^9,14,16,33 However, same-day appointments and reducing lead-time introduce a number of logistical difficulties limiting implementation across all clinical settings. All of these methods mitigate the financial burden on the clinics associated with no-show rates, but they fail to acknowledge the underlying patient factors leading to no-shows. Many of these factors may be expected to disproportionally affect economically disadvantaged family. These could include communication, access to public transportation, as well as parental leave and work obligations.

Looking specifically at appointment season, our study shows that patients are more likely to no-show summer appointments. These results contradict previous studies noted in the introduction. One would expect no-show rates to peak in the winter months due to inclement weather and the proximity to the holidays. During summer months, many rural agrarian communities face a significant farm workload that might require all family members to be present with little or no time to travel outside their community. The mild climate of the southeast paired with changes in community activities during the summer months may contribute to the observed discrepancy. The multivariate model was designed using winter as the reference level. Winter was found to have a higher no-show rate than spring (OR=0.79, p= 0.026). When comparing winter to summer directly in this model, no significant difference was observed. There was a higher no-show rate in the summer months when compared to winter, but this failed to reach significance (OR=1.23, p=0.052).

Recognizing contributing factors to clinical appointment adherence may be critical to improving patient outcomes. This is particularly important in the pediatric audiology population where delay in diagnosis and treatment can have a lifelong negative impact on developmental outcomes. Ideally, methods that utilize demographic information to identify individuals at-risk for no-shows will subsequently integrate systems to follow-up and verify that patients have the means necessary to attend the appointment. This could increase patient adherence and address the various barriers to care.⁵

This study has several limitations as well as areas where it can be expanded in the future. First, this study collected data regarding scheduled appointments rather than individual patients. If there was a single patient who no-showed for numerous appointments they were scored as multiple data entries. It is possible for this to introduce some bias in the no-show pool of the sample data. Additionally, it may be difficult to generalize the results of this study across other pediatric clinic populations. Audiology appointments are often quite involved and are frequently scheduled along with physician appointments on the same day. Not to mention, the audiometric testing might be intimidating for many young children and parents. Finally, the results in this study do not address same day or “work-in” appointments. These appointments are often scheduled while the patient is already in the clinic seeing another provider and would inflate the number of patients who did show for appointments. In terms of areas to expand on, it would be interesting to study family factors that may be important in pediatric populations. Factors such as number/age of siblings, education status of siblings, or occupational status of parents.

Conclusion:

In a pediatric population with audiology appointments, sociodemographic factors including patient age, Black/African American race, and Medicaid insurance status were associated with increase rates of no-show appointments. A thorough evaluation of scheduling and patient communication strategies is needed to improve appointment adherence and, subsequently, hearing outcomes. Further research is needed to identify and develop systems that promote appointment adherence by recognizing these factors and integrating avenues to bypass barriers to care.