What Is “Lost to Follow-Up” in Mammography?

Introduction

Lost to follow-up (LTFU) in mammography is the operational state in which a patient has not completed recommended additional imaging, biopsy, or short-interval surveillance after a mammogram, and the facility has exhausted reasonable, documented outreach attempts so that outcome information can no longer be obtained and the patient can no longer be re-engaged through standard channels. This state applies equally to patients assessed as Incomplete (Need additional imaging evaluation), Suspicious, Highly Suggestive of Malignancy, or Probably Benign. It is distinct from merely being overdue. The distinction is not semantic. It determines whether a case remains intervenable, whether the Mammography Quality Standards Act (MQSA) medical outcomes audit can classify the examination as a true positive or false positive, and whether navigation resources can still convert the recommendation into a completed diagnostic loop.

The familiar assumption is that any unresolved positive assessment simply sits in an "incomplete" or "overdue" bucket until it ages out of the tracking system. The unexpected distinction is that overdue and LTFU occupy sequential positions on the same pathway and possess different intervention capacity. Overdue status still permits active outreach because contact information is presumed current and the recommendation interval has only recently expired. LTFU status is reached only after those efforts have failed or been abandoned. Facilities that collapse the two states lose the ability to measure conversion rates, allocate navigator effort, and produce credible audit metrics.

MQSA Requirements and the Audit Data Gap

Under the MQSA regulations at 21 CFR 900.12(f), every certified facility must establish and maintain a system that collects and reviews outcome data for all mammograms, including follow-up on the disposition of all positive mammograms and correlation of pathology results with the interpreting physician's findings. Analysis must occur both individually for each interpreting physician and collectively for the facility. Amended regulations effective September 10, 2024, further require calculation of positive predictive value, cancer detection rate, and recall rate. [1]

The Food and Drug Administration (FDA) defines positive mammograms for audit purposes as those carrying a final assessment of Suspicious or Highly Suggestive of Malignancy. Incomplete assessments are stated in the official MQSA lexicon as "Incomplete: Need additional imaging evaluation" or "Incomplete: Need prior mammograms for comparison." The assessment "Probably Benign" carries a management recommendation for short-interval surveillance, most commonly at 6 months (sometimes stated as 4-6 months). The American College of Radiology (ACR) Breast Imaging Reporting and Data System(r) (BI-RADS(r)) maps these FDA assessment statements to numeric categories (category 0 for Incomplete, category 3 for Probably Benign, category 4 for Suspicious, category 5 for Highly Suggestive of Malignancy). ACR guidance asserts that a meaningful screening audit must also treat Incomplete assessments as positive because failure to track their resolution systematically undercounts the abnormal interpretation rate and subsequent diagnostic yield. [2]

When recommended additional imaging, tissue sampling, or short-interval surveillance does not occur and the case never reaches a known outcome, the examination cannot be classified as a true positive or false positive. The denominator of positive assessments therefore includes unresolved cases. Simulation work demonstrates that lower rates of pathology ascertainment increase the probability that a facility or individual radiologist will fall below established performance benchmarks for cancer detection rate. Incomplete ascertainment therefore distorts sensitivity, specificity, positive predictive value, and cancer detection rate by construction. [6]

Operational Definition and Time Thresholds

A patient becomes overdue when the interpreting physician has issued a recommendation that requires action and the recommended interval has passed without completion, yet the case remains active in the tracking system. This includes:

• Additional imaging after an Incomplete assessment (Need additional imaging evaluation),
• Tissue sampling after a Suspicious or Highly Suggestive of Malignancy assessment, and
• Short-interval surveillance imaging after a Probably Benign assessment.

Validated electronic triggers applied to large electronic-health-record cohorts flag these delays at practical thresholds of more than 60 days for Incomplete, Suspicious, or Highly Suggestive assessments and more than 7 months for Probably Benign assessments. At this stage the recommendation is still known, contact information is presumed current, and outreach can still convert the status to completed. [3]

LTFU is the later state in which the recommended action has not occurred after the overdue window, documented outreach attempts have failed to produce resolution, or the patient is no longer reachable within the health system after a longer defined period. Patients assessed as Probably Benign who fail to return for the recommended 6-month (or 4-6 month) examination after exhaustive outreach meet the definition of LTFU exactly as patients with Incomplete or Suspicious assessments do. Systematic reviews of follow-up after abnormal mammography document that investigators operationalize this state in several ways: failure to attend a scheduled appointment, failure to complete any follow-up within a fixed interval such as 3 or 6 months, or failure to reach diagnostic resolution. Rates of non-attendance within 3 months range from 7.2 percent to 33 percent across studies; rates at 6 months range from 27.3 percent to 71.6 percent. Observational work further suggests that the most common patient-level reasons include lack of time and the perception that personal health is already good. [4]

The operational difference is therefore temporal and systemic. Overdue remains a reversible, system-visible status that still permits intervention. LTFU is the status that remains after intervention capacity is exhausted or abandoned. Practices that keep these two statuses distinct can measure the conversion rate from overdue to completed and can identify the residual volume of unresolved cases-including unresolved Probably Benign short-interval cases-that enter the annual audit as data gaps.

Tracking Architecture and Closed-Loop Documentation

Effective tracking systems surface the overdue cohort in real time so that outreach can occur while contact remains feasible. Electronic triggers that flag delays at the 60-day and 7-month thresholds have demonstrated positive predictive values around 71 percent when applied to large cohorts, identifying hundreds of true delays among thousands of abnormal examinations. Once flagged, the system documents each outreach attempt, the outcome of that attempt, and the final disposition. Practices that maintain this layered status architecture can calculate three independent rates: the proportion of positive assessments that become overdue, the conversion rate from overdue to completed, and the residual LTFU rate after outreach. These rates feed directly into the MQSA audit and allow targeted allocation of navigator effort across both the rapid-action pathway (Incomplete, Suspicious, Highly Suggestive of Malignancy) and the short-interval surveillance pathway (Probably Benign). [3]

Patient navigation interventions increase both screening uptake and follow-up after abnormal results. A systematic review and meta-analysis of randomized trials by Nelson and colleagues reported a risk ratio of 1.23 for completion of follow-up after abnormal mammography among women offered navigation compared with usual care. The mechanism is closed-loop documentation: the recommendation is identified, the patient and referring clinician are notified, completion is confirmed, and the outcome is recorded so the loop is formally closed. Open-loop records stop at the recommendation itself. In the absence of systematic tracking to resolution, the same recommendation that appears complete on the radiology report remains unresolved in the patient's longitudinal record. [8]

Closed-loop imaging programs that index radiology reports containing follow-up recommendations, assign nurse coordinators to chart review and outreach, and document resolution have increased adherence rates for actionable findings and reduced the volume of unresolved cases. These programs treat overdue status as the trigger for active intervention and reserve LTFU status for cases that remain unresolved after defined outreach protocols (typically multiple telephone attempts followed by written notification, certified mail, and primary-care notification). The same escalation logic applies to Probably Benign short-interval cases. Open-loop environments lack this escalation path; the recommendation is issued and the case simply ages into permanent non-resolution. [7]

Patient Consequence

Longer intervals between an abnormal screening result and diagnostic resolution are associated with higher odds of larger tumor size and lymph-node involvement once cancer is diagnosed. The Population-based Research Optimizing Screening through Personalized Regimens (PROSPR) Consortium systematic review concludes that risk for poorer cancer outcomes rises with prolonged wait times, supporting diagnostic testing as soon as feasible after a positive result, while acknowledging that evidence quality remains limited by confounding. Every case that transitions from overdue to LTFU-including cases that began as Probably Benign short-interval recommendations-therefore represents a measurable erosion of the mortality benefit that screening is intended to deliver. [5]

Counterarguments exist. Some unresolved cases reflect patient preference, relocation outside the health system, or care obtained elsewhere. Facilities that document exhaustive outreach and primary-care notification fulfill their regulatory obligation even if the ultimate disposition remains unknown. Yet the audit still carries a data gap, and the patient still carries an unresolved risk. The systems-level response is not to eliminate every LTFU case-an unrealistic goal-but to minimize the conversion from overdue to LTFU through real-time visibility, standardized outreach protocols, and closed-loop confirmation for every recommended action, including short-interval surveillance after a Probably Benign assessment.

Operational Implication and Leadership Question

Practices that maintain separate, time-stamped statuses for overdue and LTFU gain a practical set of process metrics that map directly onto the MQSA audit requirements, support focused navigator effort, and generate the documentation needed to demonstrate that the facility has fulfilled its obligation to track disposition of positive examinations. Facilities that collapse the two states into a single "incomplete" category cannot isolate the intervenable population-whether that population carries an Incomplete, Suspicious, Highly Suggestive of Malignancy, or Probably Benign assessment-and therefore cannot improve the conversion rate that actually determines patient outcome.

If the tracking system cannot yet distinguish a patient who is still reachable and overdue from a patient whose diagnostic loop has already failed, what single process change would most improve the volume of unresolved positive examinations that enter the next annual audit as permanent data gaps?

---

References (AMA 11th edition)

1. US Food and Drug Administration. Mammography Quality Standards Act regulations, 21 CFR 900.12(c) and (f). Amended requirements effective September 10, 2024. https://www.fda.gov/radiation-emitting-products/mammography-quality-standards-act-mqsa-and-mqsa-program/important-information-final-rule-amend-mammography-quality-standards-act-mqsa
2. American College of Radiology. ACR BI-RADS Atlas, 5th Edition. Follow-up and Outcome Monitoring. Reston, VA: American College of Radiology; 2013 (updated guidance reflecting 2024 MQSA amendments).
3. Murphy DR, Meyer AND, Vaghani V, et al. Electronic triggers to identify delays in follow-up of mammography: harnessing the power of big data in health care. J Am Coll Radiol. 2018;15(2):287-295. doi:10.1016/j.jacr.2017.10.001. https://www.jacr.org/article/S1546-1440(17)31272-3/abstract
4. Reece JC, Neal EFG, Nguyen P, McIntosh JG, Emery JD. Delayed or failure to follow-up abnormal breast cancer screening mammograms in primary care: a systematic review. BMC Cancer. 2021;21(1):373. doi:10.1186/s12885-021-08100-3. https://bmccancer.biomedcentral.com/articles/10.1186/s12885-021-08100-3
5. Doubeni CA, Gabler NB, Wheeler CM, et al. Timely follow-up of positive cancer screening results: a systematic review and recommendations from the PROSPR Consortium. CA Cancer J Clin. 2018;68(3):199-216. doi:10.3322/caac.21452.
6. Burnside ES, Lasarev MR, Sprague BL, et al. The importance of outcomes ascertainment for accurate assessment of the mammography screening cancer detection rate: a simulation study. J Am Coll Radiol. 2024;21(3):376-386. doi:10.1016/j.jacr.2023.08.045. https://www.jacr.org/article/S1546-1440(23)00845-1/abstract
7. Lippert RD Jr. Overdue and Lost to Follow-Up as Distinct States: Operational Implications for the Medical Outcomes Audit, Tracking Architecture, and Closed-Loop Navigation. Mammologix. August 17, 2026. https://mammologix.com/mammotrack/overdue-lost-to-follow-up-distinct-states-mammography-tracking
8. Nelson HD, Cantor AG, Pappas M, Blackie K, Yu Y, Fu R. Patient Navigation Services for Breast and Cervical Cancer Screening and Follow-Up: A Meta-Analysis. JAMA Intern Med. 2025. doi:10.1001/jamainternmed.2025.XXXX (RR 1.23 for follow-up after abnormal mammography; 3 RCTs, n=1008).