Care has gone continuous. Operating system hasn’t

Non-communicable diseases caused at least 43 million deaths globally in 2021, around 75% of all non-pandemic deaths that year, and 82% of those premature deaths occurred in low-and middle-income countries. That single statistic captures why the shift toward continuous, preventive care is not a niche wellness trend. It is a response to how disease actually plays out over decades, not single episodes, and India’s own push toward longitudinal, consent-based health records reflects the same underlying shift.

What has not moved at the same pace is the operating system underneath it. Most healthcare organisations, including many proactive and preventive clinics leading this shift, are still built around the encounter as the basic unit of operations. Scheduling systems, electronic health records and practice management tools were designed for a model where care happens in bursts: a booking, a visit, a bill, a record. That architecture is effective when care is episodic. It becomes increasingly strained once monitoring, communication and intervention have to continue between visits, not just during them.

This is not really about any one piece of software falling short. EHRs document care and support compliance well. CRMs manage acquisition and communication well. Each does what it was designed to do. Often, the problem is not inside any one system. It lives in the seams between them, when a single patient’s journey has to move across several systems over months, and no one of them is built to hold the whole picture on its own.

There is a fairly counterintuitive dynamic worth naming here. A new biomarker, a wearable signal, a predictive risk score, all sound like automation. In practice, each one also creates a decision: is this meaningful, who should review it, does the patient need contact, did it happen. Every useful health signal creates an obligation to decide what happens next. In the US, wearable use among adults rose from 30.2% in 2020 to 41.1% in 2024, yet only 19.2% of wearable users reported actually sharing that data with a clinician. The pattern is a useful illustration of a much broader point: the industry is getting better at generating continuous health information faster than it is building mechanisms to absorb and act on it. Detection is not intervention. An order is not completion. A message sent is not a follow-up completed. A lab value in a portal is information. It becomes care once someone interprets it, reaches the patient, confirms the next step was taken, and decides what follows.

As care becomes more personalised, the number of possible pathways increases. Different patients have different testing intervals, risk thresholds, interventions, communication preferences and reassessment schedules. Personalisation creates clinical value partly by introducing variation. Operationally, that variation has to be managed, and that is worth acknowledging alongside its clinical upside, particularly at scale, where higher patient volumes make manual coordination harder to sustain.

This connects fairly directly to the business model underneath these clinics. A business paid primarily for encounters can organise much of its economics around completed visits. A six-month programme or annual membership is selling something different: continuity itself. A patient can move from lead to enrolled to onboarded to activated to engaged to retained, and each of those is a distinct operating state. A traditional appointment system usually has no reason to model all of them, but a continuity-based business depends on tracking the gaps between them.

Clinician time sits inside this picture too. In 2024, US physicians surveyed by the AMA reported working an average of 57.8 hours a week, but only 27.2 of those hours were spent in direct patient care, with the rest split between indirect care and administrative work. The proportions will vary by market, but the underlying constraint holds everywhere: a continuous-care model cannot simply respond to every additional datapoint by consuming more of a clinician’s limited attention.

It is tempting to treat this as a procurement question: add a platform, connect two systems. It may be more useful as an organisational design question, since responsibility for one patient’s journey often sits across marketing, clinical care, labs and finance, each with its own sense of what “complete” means. Integration of data does not automatically create integration of responsibility. Systems of record are designed to preserve what happened. Continuous care also needs reliable mechanisms for determining what should happen next, assigning it, following it through, and knowing when the loop is closed.

We are redesigning healthcare around continuous prevention and personalisation. The harder question, and the one worth sitting with, is whether we are redesigning the operating system required to deliver it.