Patient Leakage in Healthcare: Causes, Cost and How to Measure It

What You Need to Know

• Patient leakage can happen at four points: a referral that never becomes a visit, a caller who can't book, an established patient who stops returning, or care that goes to outside providers.
• In one academic health system, about 35% of primary care referral scheduling attempts ended in a completed specialist visit on record¹. Many completed visits may never have reached its electronic health record (EHR)¹.
• Measure each type of patient leakage from its own data. On the phone side, your call volume and missed call rate give a first estimate of how many patients you lose.

What is patient leakage, and where does it happen?

Patient leakage is when your patients leave your practice or network for care you could have provided, or go without that care. It happens at four points: referral leakage, access leakage, patient attrition and network leakage. A different team can act on each one.

TypeWhat leaksWhereWho actsReferralReferred patient seen elsewhere or not at allReferral to specialist visitReferral teamAccessCaller can't bookPhones, waits, hoursAccess centerAttritionEstablished patient stops returningBetween visitsClinic operationsNetworkCare goes to outside providersWhole care episode (claims)Network strategy

What causes patient leakage, and why don't network contracts stop it?

Patient leakage happens when getting care from you is harder than going elsewhere¹,². The appointment may be too far out, the referral may never get scheduled, or your organization may be hard to deal with¹,².

Network contracts can't stop it alone, because patients and insurers can still choose another provider³. Federal rules also limit how a health system can require its physicians to refer in-network³.

Why patients leave

In one large health system, scheduling problems, longer waits and slightly longer distances contributed to referrals with no completed visit on record¹. Differences between its clinics contributed too¹.

About one in five U.S. adults said they had switched providers in the past year, in a survey by consulting firm Accenture². Nearly 90% of those switchers said they left because the organization was "hard to do business with"². The guide to patient experience in healthcare covers how practices measure that experience.

Switchers also named clinical experience or expertise (44%) and value (10%) as reasons². Access was a reason for 13%, including inconvenient appointment hours, not getting an appointment quickly and inconvenient locations². When they chose a new provider, though, 70% of switchers cited access as a deciding factor².

Why network contracts don't stop it

Joining Medicare's Shared Savings Program didn't clearly speed up the decline in specialty-visit leakage at the most specialty-oriented accountable care organizations (ACOs)⁴. Leakage fell from 70% to 68% for ACOs that entered in 2012 and from 64% to 61% for 2013 entrants⁴. Those declines began before the ACOs joined⁴.

The Stark Law, the federal physician self-referral law, sets rules for tying physician pay to referrals for designated health services that Medicare pays for³. Its directed referral rule, 42 CFR 411.354(d)(4), allows tying that pay to referrals to a particular provider only when the arrangement meets every condition³.

The pay must be set in advance at fair market value and fit a Stark Law exception³. Neither the pay nor the arrangement itself may depend on the number or value of referrals³.

The referral requirement must be in writing, signed by the parties and limited to services within the scope of the physician's employment or contract³. It must also be reasonably necessary for the arrangement's legitimate business purposes³. Even then, the rule names three cases where the referral requirement does not apply³.

• The patient prefers another provider.
• The patient's insurer determines the provider.
• The referral is not in the patient's best medical interests in the physician's judgment.

Plans that direct referrals can also raise issues under the federal Anti-Kickback Statute and state law. Have counsel review any such plan before you adopt it.

How do you measure patient leakage?

Track each type of patient leakage as its own rate. For each one, set who counts as lost, the group you measure against, a time window and a data source. Compare each rate with your own earlier periods, measured with the same definition and window.

TypeFormulaData sourcePublished figureReferralReferrals with no in-network visit ÷ referralsEHR referrals + visits34.8% of scheduling attempts ended in a visit (one system)¹AccessBooking callers with no appointment ÷ booking callersCall records + scheduleNone publishedAttritionPatients with no visit in 18 months ÷ patients seenVisit historyUnder 19% (no prior no-shows) to almost 32% (1+ no-shows), vendor data⁵NetworkOutside visits or spend ÷ all, for attributed patientsPayer or ACO claims61% to 68% of specialty visits, Medicare ACOs, 2014 (may overstate)⁴

Each published figure describes one patient group, setting and period, so treat the closest match as a rough check, not a target.

For access leakage, use the call definitions in healthcare call center metrics. Count a caller as leaked only after your booking window closes. Set that window by visit type, since an urgent referral and a routine follow-up need different ones. Third next available appointment, the days from a request to the third open slot, measures how long patients wait in your clinics.

Referral records miss care that patients get elsewhere¹. The referral study counted only visits completed in the system's own EHR, and every out-of-network scheduling attempt in it had no recorded appointment date¹. Match referral data to claims to see where those patients went.

A referred patient who can't get through by phone counts as both referral and access leakage. Count each patient only where you first lost them.

Patients who stop returning may never call or cancel, so attrition can leave no trace in your records. Compare visit histories across a fixed window to find it.

What does patient leakage cost, and how much of it can you recover?

The cost of patient leakage is the number of patients you lose times the value of the care they would have gotten from you. Your open appointments cap how much of that you can win back.

Here's a worked example for phone leakage, with round numbers to replace with your own.

1. Monthly inbound calls: 20,000
2. Unanswered calls at 10%: 2,000
3. Scheduling-intent share of those calls at 50%: 1,000
4. Callers with no appointment within your window at 30%: 300 leaked patients per month
5. First-visit revenue at $200 per patient: $60,000 per month, or $720,000 per year before any follow-up care

To rerun it with your numbers, enter your monthly call volume and missed call rate in the Access Center Recoverable Revenue Calculator. It caps recovered appointments at your open slots and gives a rough estimate.

Referral leakage works the same way: multiply referrals issued by the share that never become visits, then by what each completed visit is worth. If first-visit value differs across service lines, give each specialty its own estimate.

Revenue lost after the visit, in billing and coding, is a separate problem for revenue integrity work.

Which type of patient leakage can Commure AI Call Center Agents reduce?

Commure AI Call Center Agents address access leakage on the phone, where patients call after hours or at a busy time and can't get through. They pick up at any hour with no hold time, schedule and reschedule in English and Spanish by your rules, and pass complex calls to your staff.

For referral leakage, Commure offers Commure Orchestrator for referrals that stall before a visit.

To see how your calls end today, ask for a complimentary call center health analysis. It covers a week of your call transcripts, reviewed under a signed business associate agreement, and breaks call volume down by type and outcome to check against your bookings.

Where should you start reducing patient leakage?

Start with access leakage, because you can measure it from call records and the schedule you already have. Then give each measure an owner on the team that can act on it. The guide to patient access covers the front-end work from first contact to payment.

1. Match call records to your schedule to measure access leakage, and set answer-speed targets with healthcare call center best practices.
2. Pull referral completion from your EHR, and report CMS50, the CMS measure for closing the referral loop, beside it as its own rate⁶. CMS50 looks at each patient's first referral made on or before October 31⁶. It counts whether the specialist's report reached the referring clinician, not completed visits⁶.
3. Track 18-month return rates for established patients, and route missed follow-ups and no-show rebooking back to scheduling.
4. Use payer or ACO claims for network leakage, and record the dates each claims feed covers.

Report all four patient leakage rates on one page, each with its window and data source.