How One Medical's operational DNA enabled rapid COVID-19 response

Executive overview

Most healthcare providers scrambled when COVID-19 hit. One Medical was able to scale testing, launch virtual care workflows, and partner with city governments within weeks. The difference wasn't luck — it was a pre-built operating system designed for exactly this kind of speed.

One Medical's membership-based, multimodal care model — combining digital health, in-person clinics, and employer relationships — meant COVID adaptations were extensions of existing infrastructure, not from-scratch builds.

The core insight: startup-speed execution at scale comes from standardizing how you work every day, not from improvising in a crisis.

How COVID hit One Medical's business

• IPO launched January 31, 2020 — the same day markets were reacting to early COVID news
• Digital health engagement surged as members flooded the platform with anxiety-driven outreach
• In-person volumes fell as shelter-in-place orders spread across markets
• Positive test rates varied dramatically: 15–20% in New York City vs. 3% in Phoenix
• Served ~500,000 members across 7,000 companies including Google and SpaceX

The operational system that made speed possible

• TOPS (One Medical Performance System) standardized operations nationally — labs, specimen vials, PPE supply chains, phlebotomy were all pre-existing
• Because workflows were standardized, the company could push new protocols to thousands of staff in one to two days
• Converted some offices into respiratory care clinics (full PPE, ER-level precautions); kept others for non-COVID patients
• Built COVID screening questionnaires into the app to triage demand before it hit providers
• Improvised where needed (buying painters suits from Home Depot) while the core system held

Multimodal care in practice during the pandemic

• Members could reach care via text, message, call, video, or in-person — including drive-up parking lot testing
• Outbound proactive outreach: contacted all members to ask about symptoms, household exposure
• Structured COVID questionnaire routed members to the right channel: scheduling, testing, or virtual visit
• 90%+ retention on both direct-to-consumer and enterprise accounts meant a stable, known member base to proactively manage
• In normal times: 7 average engagements per member per year; 47% monthly active use rate

Understanding COVID testing

• PCR testing is the primary tool for acute diagnosis — most accurate for detecting active infection
• Test sensitivity and specificity interact with community prevalence to determine real-world accuracy
• Example: a 90% sensitive test in a 5% prevalence community may yield only ~50% positive predictive value
• Higher prevalence (e.g. New York at 15–20%) changes the calculus — different protocols were appropriate for different cities
• IgM antibodies appear earlier in infection; can be combined with PCR to extend the detection window in the first 7–10 days
• IgG antibodies appear later and are associated with potential immunity — though duration and mutation effects remain uncertain

Workplace reentry: the Healthy Together program

• Goal: identify who is currently infectious, not just who was previously infected
• Daily screening questionnaire for employees: symptoms, household exposure, travel, temperature
• Machine learning models showed symptom questionnaire responses are highly correlated with PCR-positive results
• Temperature checks alone miss asymptomatic carriers — temperature should be taken at home, not at the door
• Employers receive aggregate risk dashboards: e.g. "4,900 employees: no symptoms; 100: symptomatic or intermediate"
• Full testing workflow embedded in the app: screen → test → results → provider consultation

Future planning and strategic alignment

• Scenario planning ran in parallel with day-to-day crisis response from the earliest weeks
• Three-horizon thinking: acute outbreak → antivirals → vaccine distribution
• Planning framework: strategic alignment (cross-functional scenario teams) → improvement and innovation (lean, agile, design thinking) → active daily management (standards, training, field observation)
• Technology stack built with modern APIs and agile teams — allows rapid repositioning
• New proof points emerged: partnership with Montage Hotels for guest screening; JAMA Network Open paper showing 45% reduction in employer healthcare costs

The broken US healthcare baseline

• US spends 18% of GDP on healthcare; over a third considered waste
• Only 5–7% of the premium dollar goes to primary care (vs. ~14% in OECD nations)
• Average wait time for a family practitioner: 29 days
• Average family premium: $20,000
• 50% of US family practitioners show burnout symptoms — driven by fee-for-service incentives that reward volume over outcomes
• One Medical's response: salaried providers, proprietary EHR with 44% less administrative burden, insurance acceptance, and digital interfaces to hospital systems and specialists