Segmenting outpatient appointment behavior: a data-driven approach to understanding no-shows, delays, and walk-ins in...

Outpatient appointment disruptions-particularly no-shows, late arrivals, and walk-ins-undermine clinic efficiency and equity in access to care. Despite decades of research, predictors of attendance behavior remain poorly understood, especially in respiratory care settings.

We conducted a cross-sectional study of 520 patients attending a pulmonology outpatient clinic in Turkey over 12 weekdays. Patients were classified as on-time attendees, late attendees, no-shows, or walk-in visitors. Data on sociodemographic, clinical, logistical, and behavioral factors were collected via structured surveys administered in person or, for no-shows, by phone; some variables (e.g. employment and insurance) had substantial missingness. Logistic regression identified predictors of non-attendance. An unsupervised clustering analysis (Partitioning Around Medoids using Gower distance) was used to identify patient subgroups with distinct attendance patterns.

Among 520 patients, 386 (74.2%) arrived on time, 64 (12.3%) were late, 48 (9.2%) missed appointments, and 22 (4.2%) were walk-ins. On-time attendees were older and more likely to have received appointment reminders. No-shows were younger, more educated, and less likely to receive reminders. The leading causes of lateness were transportation issues (45.9%) and difficulty locating the clinic (31.1%). In multivariable analysis, younger age (odds ratio [OR] = 0.964; 95% confidence interval [CI] = 0.938-0.992; P  = .010) and absence of reminders (OR = 4.275; 95% CI = 2.013-9.081; P  < .001) were independently associated with non-attendance. Clustering analysis identified three phenotypes: "Older Dependent" (older, comorbid, low socioeconomic status), "Young Autonomous" (younger, high socioeconomic status, high no-show rate), and "Access-Challenged" (employed, low insurance coverage, high lateness).

Appointment reminders and age are strong, actionable predictors of attendance. Behavioral segmentation revealed latent patient profiles with distinct needs. Personalized scheduling strategies-especially targeted reminders and flexible systems-may improve outpatient efficiency and reduce inequities. These findings support the integration of behavioral clustering into appointment optimization frameworks.Keywords: outpatient appointments; no-show behavior; appointment reminders.