Does online consultation end the 8am rush?
NHS England · Cloud Based Telephony & OC submissions · February 2026 · per working day
Design: Cross-sectional comparison of 4,306 practices with both telephony (CBT) and OC data, grouped into tertiles by current OC rate (Feb 2026, per 1,000 patients per working day).
Low OC (n=1,435) — <2.1/1k/wd · median 8 submissions/day · essentially phone-first.
Mid OC (n=1,436) — 2.1–9.2/1k/wd · median 43/day · OC as secondary channel.
High OC (n=1,435) — ≥9.3/1k/wd · median 144/day · heavily OC-dependent.
Data: CBT001 = inbound calls. CBT003 = answered. All rates per 1,000 patients per working day (Feb 2026: 20 working days).
1. Calls in by time of day
Inbound calls per 1,000 patients per working day
CBT001 · February 2026 · 8am–6pm
Reading: Low-OC practices receive a large 8–10am phone spike (~11/1k/wd) while high-OC practices receive roughly half (~5.4/1k/wd). The gradient between tertiles is smooth and consistent across the day.
2. OC submissions by time of day
OC submissions per 1,000 patients per working day
February 2026 · 8am–6pm
Reading: OC has its own 8–10am rush: 5.3/1k/wd for high-OC practices, peaking in the same window as the phone spike it displaces. The OC spike is comparable in magnitude to the phone calls it replaces. Low-OC practices have almost no OC volume. The rush has not been eliminated — it has changed channel.
3. Total demand at 8–10am
8–10am demand breakdown by tertile
Stacked: answered + unanswered calls + OC submissions · per 1,000/wd · Feb 2026
Reading: If we count all inbound calls (answered and unanswered) plus OC, the totals look comparable: 11.3, 9.9, 10.7/1k/wd. But unanswered calls don’t generate work — they bounce off. If we count only the demand that actually hits the practice (answered calls + OC submissions), the picture changes: Low OC 6.0, Mid 6.1, High 8.4/1k/wd. High-OC practices receive 40% more actionable morning demand. And 33% of their day’s work lands in the 8–10am window, compared to 29% for Low OC. OC does not end the 8am rush — it makes it worse.
4. Total daily demand (8am–6pm)
Full-day demand breakdown by tertile
Stacked: OC submissions + answered calls + unanswered calls · per 1,000/wd · Feb 2026 · 8am–6pm
Reading: Total inbound contacts (including unanswered calls) are similar: 32.6, 31.0, 34.3/1k/wd. But actionable demand (answered calls + OC) is higher in High OC: 25.5 vs 20.7/1k/wd. Every OC submission requires clinical processing — reading, triaging, responding — that almost certainly doesn’t appear in appointment data.
5. Calls answered by time of day
Calls answered per 1,000 patients per working day
CBT003 · February 2026 · 8am–6pm
Reading: High-OC practices receive fewer calls, so you might expect them to answer the same number from a shorter queue. They don’t — they answer fewer calls too. The shorter queue has not translated into better phone access.
6. Answer rate by time of day
Percentage of inbound calls answered
February 2026 · 8am–6pm
Reading: At 8–10am, high-OC practices answer a higher share of calls (58% vs 52%) because the queue is shorter. But from 10am onwards, low-OC practices answer 7–9 percentage points more (64–70% vs 57–63%). The rush-hour advantage does not carry through the day.
7. The full picture
Daily totals per 1,000 registered patients per working day
8am–6pm · February 2026
| Calls in | Answered | Unans. | OC | Total |
|---|
| Low OC | 31.6 | 19.7 | 12.0 | 1.0 | 32.6 |
| Mid OC | 26.3 | 16.0 | 10.2 | 4.8 | 31.0 |
| High OC | 21.8 | 13.1 | 8.7 | 12.4 | 34.3 |
Key finding: Total inbound contacts are similar across tertiles (~31–34/1k/wd), but actionable demand (answered calls + OC) is higher in High OC (25.5 vs 20.7/1k/wd). High-OC practices receive ~10 fewer phone calls per 1k/wd, but ~11 more OC submissions — and every submission requires clinical work. OC has not created spare capacity.
8. What kind of OC demand?
OC submissions by type — clinical, administrative, other/unknown
February 2026 · 4,686 practices with both CBT and OC data
Reading: In the lowest-OC tertile, over a third of submissions are administrative (38%). From October 2025, the GP contract required practices to keep their OC tool open during core hours for “non-urgent appointment requests, medication queries and admin requests” — many low-OC practices may have switched it on to meet this minimum and the traffic they attract is skewed towards those categories. In the highest-OC tertile, two-thirds of submissions are clinical (67%). High-OC practices are not just handling more admin online — they are receiving proportionally more demand that requires clinical review.
Does OC end the 8am rush? No — it makes it worse. When we count only the demand that actually hits the practice (answered calls + OC submissions), high-OC practices receive 8.4/1k/wd in the 8–10am window versus 6.0 for Low OC — 40% more. And 33% of their day’s actionable demand is concentrated in that two-hour slot, compared to 29% for Low OC. The phone queue is shorter at 8am, which briefly improves the answer rate, but by 10am low-OC practices are answering a higher share of calls.
Any sign of freed-up capacity? Not in these data. Actionable demand is actually higher in high-OC practices (25.5 vs 20.7/1k/wd) because every OC submission generates clinical work — reading, triaging, responding — that almost certainly does not appear in appointment data. If OC were freeing up capacity, we would expect fewer total contacts to process, or more calls answered from a shorter queue, or a higher answer rate outside the rush. None of these are observed. The underlying constraint is clinical capacity: the same patients need triaging and in many cases seeing, regardless of which channel they used.