Large-Scale No-Show Patterns and Distributions for Clinic Operational Research

Patient no-shows for scheduled primary care appointments are common. Unused appointment slots reduce patient quality of care, access to services and provider productivity while increasing loss to follow-up and medical costs. This paper describes patterns of no-show variation by patient age, gender, appointment age, and type of appointment request for six individual service lines in the United States Veterans Health Administration (VHA). This retrospective observational descriptive project examined 25,050,479 VHA appointments contained in individual-level records for eight years (FY07-FY14) for 555,183 patients. Multifactor analysis of variance (ANOVA) was performed, with no-show rate as the dependent variable, and gender, age group, appointment age, new patient status, and service line as factors. The analyses revealed that males had higher no-show rates than females to age 65, at which point males and females exhibited similar rates. The average no-show rates decreased with age until 75–79, whereupon rates increased. As appointment age increased, males and new patients had increasing no-show rates. Younger patients are especially prone to no-show as appointment age increases. These findings provide novel information to healthcare practitioners and management scientists to more accurately characterize no-show and attendance rates and the impact of certain patient factors. Future general population data could determine whether findings from VHA data generalize to others.


Introduction
No-show patient appointments have been defined as "patients who neither kept nor cancelled scheduled appointments" [1]. Although the documented rates of missed appointments may vary somewhat between countries, health care systems, and clinical settings [2], appointment-breaking behaviors constitute a widespread, global issue [2][3][4]. No-show rates have been shown to range from 15%-30% in general medicine clinics and urban community centers [5,6]. It has also been reported that no-show rates can reach as high as 50% in primary care [7].
Tampa, and Houston. The VHA outpatient appointments were grouped into stop codes and then aggregated into six service lines: Primary Care; Mental Health; Specialty Medicine; Rehabilitation; Surgery; and Other. For a general description of service line departments, please see Table A1 in the appendix.
The available data fields included a de-identified patient ID number, patient age at the time of the appointment, gender, the date the appointment was entered into the system, the date and time for which the appointment was scheduled, and the final appointment status (complete or incomplete). Additionally patients were classified into new or already established for a given appointment. New patients were those who did not complete an appointment within a given clinic type during the prior 24 months. Some patients were active in the VA outpatient system for all years studied, while others were active for only a subset of those years. The appointment age was determined by calculating the number of days between the date the appointment was created and the date the appointment was completed. The final data set included 555,183 patients, who scheduled 25,050,479 appointments.

Data Analyses and Methods
Multifactor analysis of variance (ANOVA) was performed, with no-show rate as the dependent variable and the following factors; A. New versus Established patients; B. Appointment Age Groups; C. Patient Age Groups; D. Gender; E. Service Line. Pairwise comparisons were performed using 95% Tukey intervals (other interval types gave similar results). All interactions of order two were included to uncover the significance of the main effects on the no-show rate. This allowed for a more detailed look at the interactions between gender, appointment and patient age, and whether the patient was new or established. Service line was included in the analysis to factor in differences in scheduling.

Results and Discussion
The following section describes the analysis and discusses the results of the subsequent ANOVA. Table 1 shows that the main effects and all the interactions of order 2 are significant. Note that the interaction between Appointment Age Group and Gender has the smallest significance as indicated by the F-Ratio in Table 1. All f -ratios are based on the residual mean square error.

Gender and Age Frequencies
The VHA population was predominately male (91.47%) and between 60 and 70 years of age or older (29.32% of the total population). Table 2 provides a breakdown of 14 age groupings compared to the total population. This sample contains about 10 times more males than females. The highest female appointment frequency occurs at age 45 to 64 compared to males at age 55 to 69. This pattern reflects recent national trends of more women in the military.  Figure 1 shows the aggregated no-show data trends. The overall average no-show rates decrease with age until 75-79, when they increase slightly. Males have higher no-show rates than females until age 65, when males and females exhibit similar rates. Figures 2 and 3 compare the overall pattern segmented by service line, gender, and age. The Medical, Primary Care, and Surgery Service Lines have patterns similar to the overall results in Figure 1. Interestingly, Mental Health and Rehabilitation reveal females above age 74 with higher than overall expected no-show rates. While this observation could be influenced by relatively small numbers (see Table 2), it is an area for further study. Appointment frequencies are broken down, in Table 3, to reflect the total number of appointments for each service line by gender and age grouping.

Appointment Age
Appointment age is defined as the difference between the date an appointment was scheduled and the future pending appointment date. This shows "how far in advance" an appointment is created or made. Consistent with past research, no-show rates increase as appointment age increases [32,33]. While no-show rates for males were generally higher than females, males and females tended to have similar rates with respect to appointment age, shown in Figure 4. The appointment frequencies, in Table 4, show that the majority of the total appointments (63.6%) occurred between two and 65 days of lead time. There were also a large number of same day appointments. This holds true across all service line, shown in Table 5.                 New patients are defined as those who did not have a completed appointment within a single service line clinic during the prior 24 months. Figure 5 shows the no-show rates for both new and established patients based on their appointment age. While there was only a slight difference found between rates for same day appointments, nevertheless there were significant differences between new and established patients across all appointment age groups. It is hypothesized that this finding reflects a new patient's desire to address their clinical needs quickly. This data suggests clinic managers and practices should be particularly attentive to their new patient waiting times.

New versus Established Patients
New patients are defined as those who did not have a completed appointment within a single service line clinic during the prior 24 months. Figure 5 shows the no-show rates for both new and established patients based on their appointment age. While there was only a slight difference found between rates for same day appointments, nevertheless there were significant differences between new and established patients across all appointment age groups. It is hypothesized that this finding reflects a new patient's desire to address their clinical needs quickly. This data suggests clinic managers and practices should be particularly attentive to their new patient waiting times.  Figure 6 shows an interaction plot of patient age and appointment age. Each line represents appointments that fall into one of nine progressively longer appointment age time groups displayed by patient age. While overall no-show rates increase with appointment age for all patient age groups, the relative impact as shown by the shape of the lines, are different. Same Day appointments generate a relatively constant no-show rate across all patient age groups. However, as appointment age increases, younger patient's no-show rates dramatically increase compared to older patients. In addition, appointment age of even one to eight days is disproportionally higher in younger patients compared to older patients. As patient age increases, the overall pattern seen in Figures 1-3 emerges. This observation suggests that managers may consider confirming the intention to keep appointments especially for young patients with longer appointment ages.   Figure 6 shows an interaction plot of patient age and appointment age. Each line represents appointments that fall into one of nine progressively longer appointment age time groups displayed by patient age. While overall no-show rates increase with appointment age for all patient age groups, the relative impact as shown by the shape of the lines, are different. Same Day appointments generate a relatively constant no-show rate across all patient age groups. However, as appointment age increases, younger patient's no-show rates dramatically increase compared to older patients. In addition, appointment age of even one to eight days is disproportionally higher in younger patients compared to older patients. As patient age increases, the overall pattern seen in Figures 1-3 emerges. This observation suggests that managers may consider confirming the intention to keep appointments especially for young patients with longer appointment ages.

New versus Established Patients
New patients are defined as those who did not have a completed appointment within a single service line clinic during the prior 24 months. Figure 5 shows the no-show rates for both new and established patients based on their appointment age. While there was only a slight difference found between rates for same day appointments, nevertheless there were significant differences between new and established patients across all appointment age groups. It is hypothesized that this finding reflects a new patient's desire to address their clinical needs quickly. This data suggests clinic managers and practices should be particularly attentive to their new patient waiting times.  Figure 6 shows an interaction plot of patient age and appointment age. Each line represents appointments that fall into one of nine progressively longer appointment age time groups displayed by patient age. While overall no-show rates increase with appointment age for all patient age groups, the relative impact as shown by the shape of the lines, are different. Same Day appointments generate a relatively constant no-show rate across all patient age groups. However, as appointment age increases, younger patient's no-show rates dramatically increase compared to older patients. In addition, appointment age of even one to eight days is disproportionally higher in younger patients compared to older patients. As patient age increases, the overall pattern seen in Figures 1-3 emerges. This observation suggests that managers may consider confirming the intention to keep appointments especially for young patients with longer appointment ages.

Conclusions
This paper describes the variation of no-show rates with patient age, gender, appointment age, and type of request within six individual service line of the United States VHA. The analyses revealed that males had higher no-show rates than females to age 65 where males and females exhibited similar rates. Average no-show rates decrease with age until age 75-79, whereupon they increase. No-show rates increase as appointment age lengthens for all age groups. Younger patients are especially prone to no-show as appointment age increases. New patients no-show at higher rates than established patients, especially beyond 36 days of lead time. These findings suggest particular attention to female patients over age 74 in Mental Health and Rehabilitation may be warranted.
This data has several limitations. The VA population does not map directly to the general United States patient population due to the greater percentages of older, male Veterans and the fact that female Veterans tend to be younger than male Veterans [34]. This data is consistent with the overall VA population median age of 57 years and approximately 90% male. Likewise, the age distribution for females is skewed more heavily towards lower age groupings as in the overall VA population [29]. Further analyses may determine if the findings are present in a non-VA population. While this data is available by service line or type, the study did not include diagnosis-specific information. For that reason, there may be additional diagnosis-related factors influencing patient appointment attendance behaviors that influence these results.
Many additional factors have been associated with nonattendance. These factors include a patient's race and ethnicity, socio-economic status, marital status, beliefs about their symptoms, source of illness, and severity of the patient's condition [16,[25][26][27]35,36]. Further, no-shows have been associated with structural barriers, such as distance to the clinic and lack of transportation [25,37].
Additionally, patient no-shows have been shown to vary by physician characteristics, patient-physician interaction, clinic access, administrative processes, and environmental factors including team communication and on-time appointments [25]. VHA data reveals no-show variation by geographical region and rural and urban settings. To best predict and minimize no-show rates, the influence of these additional factors may be important to understand and manage.
These findings from VHA longitudinal data allow understanding of no-shows from a large and statistically significant multi-year data set with little sampling bias. Practitioners working in the areas of operational research may find the results useful in order to more accurately characterize no-show and frequent attendance rates, and patient factor interactions. As a result of this study, clinicians and managers may wish to focus special attention on young male patients, new patients, and females over 74 in Mental Health and Rehabilitation. Computer and analytical modeling, as well as scheduling system re-engineering, may use this information to answer important questions regarding patient appointment behavior predictions and profiles. Future examination of data from the general population is needed to determine if the findings are generalizable beyond this population.
Jerrold H. May, Robert J. Monte, Keri L. Rodriguez, Youxu C. Tjader, and Dominic L. Vargas edited and approved the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.

Disclaimer:
The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States Government.