Telephone access in healthcare is easy to discuss and surprisingly difficult to quantify. No U.S. federal series counts every call seeking an appointment, reporting a symptom, requesting a refill, or arriving after a clinic closes. A missed-call percentage from one practice cannot fill that gap. The national evidence is instead a mosaic: CDC counts of encounters, patient-reported access measures, administrative standards, privacy guidance, and trials of specific communication interventions.
That mosaic supports a consequential but bounded finding. The telephone remains a route into care and a tool for appointment continuity, while after-hours contacts can involve clinical risk. Yet neither encounter volume nor a reminder trial tells us how many calls a typical medical office receives—or how quickly it should answer them.
Start with care volume, not an invented call total
The CDC’s National Center for Health Statistics reports 155.4 million emergency-department visits in the United States in 2022, equivalent to 47 visits per 100 people. Of those ED encounters, 18.6 million resulted in hospital admission. These are visits, not callers, and one person may account for more than one visit. The estimates come from the National Hospital Ambulatory Medical Care Survey and carry sampling error. They nevertheless establish the scale of unscheduled acute-care use during a defined year and geography.
The CDC’s ambulatory-care survey program also covers visits to office-based physicians and health centers through the National Ambulatory Medical Care Survey. Those data describe sampled encounters and patient/visit characteristics. They do not reveal whether the patient booked by phone, portal, referral, or in person, and they do not count unsuccessful attempts to obtain care. Treating visit totals as “call demand” would therefore be a category error.
The same caution applies after hours. Emergency-department volume includes urgent and nonurgent presentations, ambulance arrivals, referrals, repeat visits, and cases for which an office call would have been inappropriate. It is evidence of substantial round-the-clock healthcare demand—not evidence that better telephone answering would have prevented a specified fraction of visits.
Patient access measures include communication, but not switchboard speed
AHRQ’s CAHPS Clinician & Group Survey offers a more direct view of access as patients experience it. The standardized 3.0 adult instrument asks, over the prior six months, how often a patient obtained urgent care as soon as needed, got appointments when needed, received answers to medical questions during regular office hours, and received answers after regular office hours.
The unit is a respondent’s reported experience, typically expressed as a distribution across “never,” “sometimes,” “usually,” and “always.” The recall period is six months. Survey sponsors can use different modes and case mixes, and AHRQ advises appropriate adjustment when comparing organizations. These questions are broader than telephone performance: an answer could arrive by telephone or another channel, and “as soon as needed” is not the same as answering within a fixed number of seconds.
That design is useful precisely because it measures an outcome patients recognize rather than a narrow telecom event. A call can be answered immediately but fail to produce a timely clinical response. Conversely, a well-designed callback may resolve the need even though the initial contact was not completed in one conversation.
Appointment communication has experimental evidence
Reminder studies provide stronger causal evidence for a narrower question: whether a specified intervention changes attendance in a defined patient cohort.
In a U.S. outpatient study published in The American Journal of Medicine, Parikh and colleagues evaluated an automated reminder system across 29,400 scheduled appointments at an academic multispecialty practice. The study abstract reports a 23.1% no-show rate without a reminder and 17.3% with an automated reminder. The unit is scheduled appointments, not unique patients; the outcome is attendance, not health status. Because the analysis was conducted in one organization and was not a national random sample, its 5.8-percentage-point difference should not be promoted as a universal effect.
A Geneva academic-hospital investigation by Perron and colleagues ran two randomized controlled trials comparing text-message and telephone reminders. Across 2,123 included patients in the two trials, both active reminder modes improved attendance relative to no reminder, while text messaging was less costly under that institution’s costing assumptions. Geography, health system, language mix, workflow, and 2010-era technology constrain transfer to contemporary U.S. settings. The trial does, however, show why channel comparisons should use a no-reminder or current-practice control rather than assume every attempted contact is effective.
Shah and colleagues conducted a randomized pediatric clinic study involving 1,854 urban children and adolescents with scheduled visits. Telephone and text-message approaches were tested in a safety-net setting. This cohort matters: access barriers, prepaid-phone turnover, caregiver responsibilities, and pediatric scheduling differ from commercially insured adult specialty care. The study’s results can inform hypotheses, not supply a benchmark for all practices.
Across these studies, “reminder” is not a uniform treatment. Timing, number validity, live versus automated delivery, message language, confirmation options, and whether the message actually reached the intended person all influence exposure. A reminder attempt is not necessarily a completed communication.
After-hours contact is a clinical pathway
A Cochrane review by Bunn and colleagues synthesized trials of telephone consultation and triage in general and out-of-hours care. Although a systematic review rather than a new national dataset, it is important for interpretation: telephone management can alter immediate service use, but the included studies differed in personnel, protocols, populations, and outcomes, and evidence about safety was limited. The lesson is not that every call should be “deflected.” It is that after-hours response is part of clinical routing, where escalation rules and professional scope matter.
Privacy is another part of access. HHS states that the HIPAA Privacy Rule permits providers to communicate appointment reminders and may permit a message with family members or on voicemail, but providers should use reasonable safeguards and limit disclosed information. HHS also explains disclosures permitted for treatment, payment, and healthcare operations. Those permissions are not instructions to include diagnosis or sensitive detail in every message; patient requests for confidential communications and applicable state law still matter.
CMS communication rules provide additional context for organizations participating in Medicare markets. The Medicare Communications and Marketing Guidelines distinguish informational communications from marketing and set requirements for plans and their representatives. They should not be generalized into a universal script for physician offices, and they do not create a single national healthcare call-answer standard.
What can responsibly be concluded
The evidence supports four conclusions. First, unscheduled care demand is large and continuous, but ED visits are not a proxy for medical-office calls. Second, validated patient-experience questions treat timely answers—during and after office hours—as part of access. Third, controlled studies show reminders can improve attendance in particular settings, although effect size varies by cohort and delivery. Fourth, after-hours communication may require triage and escalation rather than administrative message capture alone.
Important gaps remain. Federal encounter surveys omit failed calls and many channel details. Reminder trials can become outdated as mobile use and portals change. Survey responses involve recall and nonresponse. None of the cited sources establishes an ideal pickup time, staffing ratio, or universal no-show reduction. The most evidence-consistent approach is to measure the complete pathway locally: contact attempts, successful connections, response time by urgency, appointment outcome, escalation, and patient experience—while keeping administrative access distinct from clinical advice and emergency care.