Healthcare works best when providers know patients' histories—previous diagnoses, treatments tried, medications taken, allergies, preferences, and the full context of health situations. But healthcare is fragmented across providers, facilities, and systems that often don't share information. Patients repeat their histories endlessly. Critical information falls through cracks. The continuity of care that good health outcomes require proves elusive in disconnected systems.
Why Continuity Matters
>Continuity of care improves health outcomes. Patients with consistent primary care providers have lower mortality, fewer hospitalizations, and better management of chronic conditions. The relationship between patient and provider—built over time through accumulated understanding—produces better care than episodic encounters with strangers.
>Information continuity prevents errors. Knowing medication history prevents harmful interactions. Knowing allergy history prevents dangerous exposures. Knowing what's already been tried prevents repeating ineffective treatments. When providers know patients' histories, they make better decisions.
>Efficiency improves with continuity. Tests aren't repeated unnecessarily when results are accessible. Consultations don't require reconstructing histories from scratch. Time spent gathering information can instead be spent providing care.
Fragmentation Realities
>Despite knowing continuity matters, healthcare remains fragmented. Different providers use different systems. Systems in different facilities don't talk to each other. Even within facilities, different departments may not share information effectively. The patient is often the only connection between disconnected pieces of their care.
>Patients with complex needs interact with many providers—specialists, primary care, emergency, rehabilitation, home care, pharmacy. Each provider sees one piece of the picture. Coordinating across providers is left to patients themselves, who may lack the knowledge or capacity for this coordination.
>Transitions are particularly risky. Moving from hospital to home care, from one province to another, or from pediatric to adult services all involve handoffs where information is lost. The patient's needs don't pause for administrative transitions, but care often does.
>Mental health and physical health often remain siloed despite their interconnection. A psychiatrist may not know a patient's medical conditions; a cardiologist may not know about depression affecting medication adherence. The whole-person perspective that good care requires is lost in specialty divisions.
Electronic Health Records
>Electronic health records (EHRs) promised to solve information continuity. Digital records that travel with patients, accessible to all providers, would eliminate lost information. The promise has been partially fulfilled at best.
>EHR implementation has been expensive and difficult. Systems purchased don't always work as promised. Training requirements are substantial. Workflow disruptions during transitions create their own problems. The benefits take time to realize while costs are immediate.
>Interoperability—systems' ability to share information with each other—remains inadequate. Different facilities use different EHR vendors whose systems don't communicate well. Provincial health information exchanges exist but don't span all providers. The fragmentation persists in digital form.
>Patient access to their own records varies. Some jurisdictions provide patient portals where individuals can see their health information. Others leave patients without access to their own data. Patient access could help with coordination, but only if patients can actually see and use their information.
Patient Burden
>When systems don't share information, patients bear the burden of coordination. They repeat their histories at every visit. They carry paper records between providers. They answer the same questions repeatedly because systems don't communicate.
>This burden falls hardest on those with complex health needs—who interact with more providers—and those with least capacity to manage it. People with cognitive disabilities, those with language barriers, those without organizational skills or support—they face the greatest coordination burden with the least capacity.
>Patients may not know what information is important to share. They may not remember medication names or dosages. They may not realize that information from one provider matters to another. The assumption that patients can coordinate their own care fails many who need care most.
Lost Information Consequences
>Information loss causes direct harm. Allergies not documented in accessible records lead to reactions. Drug interactions not caught because medication lists are incomplete cause adverse events. Diagnoses made but not communicated lead to repeated investigations.
>Delays from missing information affect care timing. Referrals sent but not received leave patients waiting without care. Results available but not accessed leave conditions untreated. The friction of information barriers slows care that should be faster.
>Lost information frustrates patients and erodes trust. Telling your story over and over, watching providers search for information you've already provided, experiencing care that ignores what you've already shared—these experiences damage the patient-provider relationship and the healthcare system's credibility.
Care Coordination Models
>Various models attempt to coordinate care across fragmented systems. Patient navigators help people move through complex care. Care coordinators manage information and communication among providers. Case managers for complex patients try to ensure coordinated care.
>These human coordination roles compensate for system failures. When systems don't share information, people must do what systems should. This adds cost and creates dependency on individuals rather than reliable systems.
>Medical homes and patient-centered medical homes attempt to establish a home base for care coordination. A primary provider or team takes responsibility for coordinating across specialists and services. This model works when medical homes have capacity and support, but not all primary care can function this way.
Indigenous Health Information
>Indigenous people's health information raises specific concerns. Historical misuse of data about Indigenous populations creates legitimate distrust. Information shared with provincial systems may be accessible to parties Indigenous communities didn't consent to. Self-determination includes governance over health information.
>Indigenous health information sovereignty—the principle that Indigenous communities should control data about their members—creates tension with mainstream health information systems. Balancing clinical continuity with Indigenous data governance rights requires approaches that mainstream systems may not accommodate.
Privacy and Sharing
>Health information privacy is legitimately important. People have interests in controlling who sees their health information. Privacy laws protect against inappropriate access. But privacy protection can also become barrier to appropriate sharing that serves patient interests.
>Consent models for health information sharing vary. Some systems require explicit consent for each sharing instance; others operate on implied consent for clinical purposes. Overly restrictive consent requirements can impede continuity; overly permissive ones can violate privacy. Finding appropriate balance requires ongoing attention.
Questions for Reflection
>Who should be responsible for coordinating care across fragmented systems—patients, primary care providers, dedicated coordinators, or systems themselves? How should this responsibility be supported?
>Should health information sharing be the default for clinical purposes, or should consent be required for each sharing instance? How do we balance continuity benefits against privacy concerns?
>How can electronic health record interoperability be achieved when vendors have commercial interests in proprietary systems? What policy interventions would help?