What the Lindsay Clancy trial reveals about disparate patient data
As I watch witness after witness testify in the Lindsay Clancy trial, I cannot help but see every place where better healthcare products might have created an opportunity to intervene before her psychiatric crisis culminated in her attempted suicide and the deaths of her three children. Her case should prompt serious examination of how healthcare systems preserve continuity when a patient’s care is distributed across multiple providers and institutions.
I am not a psychiatrist, and I am not interested in debating Lindsay's diagnosis or criminal responsibility. But I have spent more than a decade building data programs, and the evident data product failures in her case are disturbing - an enormous amount of information existed about a rapidly deteriorating patient, but it was distributed across clinicians, hospitals, health systems, prescriptions, patient statements, family observations, crisis calls, and handwritten notes. No single individual, including Lindsay herself, appears to have had a reliable, longitudinal view of her full clinical picture.
Over approximately four months, Lindsay's care touched at least five healthcare organizations and multiple outpatient and inpatient psychiatric providers. She was treated by a psychiatrist, at least two psychiatric nurse practitioners, a therapist, emergency-department clinicians, a specialized perinatal program, and staff at an inpatient psychiatric hospital. Some clinicians left their practices during her treatment, while another went on vacation during her crisis. These are ordinary personnel changes that healthcare systems must be designed to withstand. Each appointment could capture only a snapshot of her condition, while the clinically meaningful information existed in small changes occurring across time.
According to Boston Globe reporting on evidence presented at trial, Lindsay was prescribed 13 different psychiatric medications through more than 30 individual prescriptions between September 2022 and January 2023. That does not mean she was taking all 13 medications at the same time; medications were started, stopped, adjusted, and prescribed at different dosages. But reconstructing exactly what a patient was taking, when and why she was taking it, and how her symptoms changed in response required every new clinician to manually piece together information from multiple sources. It also placed an immense burden on Lindsay to reconstruct and communicate her own history each time she sought care, a responsibility made even more difficult by the psychiatric symptoms already impairing her daily functioning.
Lindsay was admitted to McLean Hospital from January 1 through January 5. She returned to outpatient treatment after her discharge and met with her psychiatrist on January 23, the day before the killings. Yet testimony reported during the trial indicates that the clinicians treating her did not consistently have the records generated by the others. Dr. Jennifer Tufts, Lindsay's outpatient psychiatrist, reportedly testified that she had not obtained the records from Lindsay's emergency-department visit or her subsequent hospitalization at McLean. Rebecca Jollotta, a psychiatric nurse practitioner who also treated Lindsay, reportedly acknowledged that she had not obtained records from Lindsay's other providers or consulted her other psychiatrist. Women & Infants Hospital of Rhode Island, where Lindsay sought admission to a specialized perinatal program, reportedly declined to admit her amid concerns that her medication burden might be contributing to her condition. According to testimony, that assessment did not reliably reach at least one clinician managing her psychiatric care.
The defense has repeatedly asked why individual clinicians did not have Lindsay sign a HIPAA authorization and obtain the missing records. It is an important question, but it misses the much bigger problem. Even if every record had been obtained, someone would still have needed to read it all, reconcile competing medication lists, organize the encounters chronologically, identify changes in symptoms, compare what Lindsay told different clinicians, incorporate observations from her family, and determine whether a meaningful pattern was emerging. That takes time, which most clinicians today do not have.
This is precisely the kind of problem AI should be helping healthcare systems solve, and this trial should be another wake-up call for the healthcare industry. The immediate capability gap is not an autonomous system diagnosing patients or making treatment decisions. It is an intelligence layer that, once a patient signs a HIPAA authorization, facilitates rapid, authorized data exchange across healthcare systems and continuously organizes the patient’s existing information into a coherent, reviewable timeline for every treating clinician. It would distinguish clinician observations from patient- and family-reported information and link every synthesized claim directly to its source record. This system could:
surface patient language indicating emerging symptoms or changes in sleep, mood, cognition, or suicidal ideation;
reconcile medications across providers and connect dosage changes to reported symptoms;
highlight recent crisis contacts, emergency visits, hospitalizations, and rejected admissions;
summarize outside assessments and communication among treating clinicians; and
flag missing records, conflicting notes, and rapid escalation for human review.
The purpose would not be to replace clinical judgment. It would be to give clinicians a better chance of exercising that judgment with the relevant information directly in front of them.
This is not merely a problem of moving data from one system into another. Healthcare has spent years working on interoperability, yet the national figures show how wide the gap remains between making information technically available and making it meaningfully usable. In 2023, 71% of U.S. hospitals reported that necessary clinical information from outside providers was routinely available electronically at the point of care. But only 42% said clinicians routinely used that information when treating patients. Just 17% reported sending care summaries to most or all outside behavioral-health providers, according to the federal government's 2023 hospital-interoperability report. The problem is even more pronounced within behavioral healthcare. In 2024, among mental-health and substance-use facilities that used only an electronic health-record system, just 48% reported electronically integrating information received from outside organizations, according to a separate federal analysis.
Availability is not synthesis. A hundred pages of records arriving in a clinician's inbox does not create a longitudinal understanding of a patient.
AI also introduces its own risks. A system could omit crucial context, exaggerate a weak pattern, reproduce errors already present in the medical record, or create automation bias. Any clinical synthesis product would need human review, direct citations to source records, transparent uncertainty, careful access controls, and rigorous testing across patient populations and clinical settings. But uncertainty about whether technology would have changed this particular outcome is not a reason to ignore what the trial is revealing: much of the information necessary to understand a patient's trajectory is already being documented and stored, but it is not being effectively assembled and presented in a form that allows an overextended clinician to comprehend it quickly.
Healthcare does not need another text box or dashboard showing clinicians more disconnected information. It needs products that perform work an individual clinician rarely has time to do manually, including turning fragmented records into an accurate, auditable, longitudinal account of what is happening to a human being. The Lindsay Clancy trial is showing us how high the stakes can be when not a single person has the complete picture.
