Skip to main content
NewNMC’s advertising guidelines for doctors are in force from 6 Oct 2026. See what to change, and get a free compliance auditNew NMC advertising rules for doctors: what to change
AI & Automation

Scoring Elective vs Urgent Leads: Why One Score Fails Joint Replacement and Cardiology

Joint-replacement enquiries and cardiac emergencies need different scoring, response targets and routing. A sample design that keeps urgent cases out of the nurture queue.

Scoring Elective vs Urgent Leads: Why One Score Fails Joint Replacement and Cardiology

A lead score is a promise about the future: this enquiry is more likely to become a patient than that one. The promise only holds if the enquiries are comparable. An orthopaedic practice receiving a call from someone whose chest pain began an hour ago and an email from someone comparing knee-replacement surgeons for next spring is not looking at two leads of different quality. It is looking at two different problems.

This article is for orthopaedic and cardiology practices, and for the marketing and operations teams who run their enquiry desks. It argues that a single score across elective and urgent demand fails in predictable ways, and sets out a sample design with separate logic, separate response targets and a safe route for anything that sounds urgent. The scoring rules below are examples to adapt, not benchmarks, and clinicians must define what counts as urgent.

If you are new to scoring, read our introduction to prioritising patient enquiries with AI first. This post goes narrower: what changes when the same inbox holds both a 12-week decision and a same-day need.

Why one score fails

A single model or rule set learns from past conversions. Past conversions in a joint-replacement practice are overwhelmingly elective: long research, several touchpoints, repeated questions about recovery and cost. The patterns that predict conversion are slow ones: multiple visits to a procedure page, a reply to a nurture message, a request for a second opinion.

Urgent enquiries have the opposite signature. They arrive once, in plain words, often at night, with no browsing history. A model trained on elective behaviour will mark them low: no repeat visits, no email engagement, no downloaded guide. The person who most needs a fast reply gets the lowest priority.

The reverse happens too. A rule that boosts anything containing "pain" or "emergency" will push a stream of elective enquiries, which mention pain as background, above the genuinely urgent. And a high-score bonus for "ready to book now" can put a keen but non-urgent patient ahead of someone whose symptoms need clinical triage.

Three things differ between the classes:

  • The decision cycle. Elective surgery decisions often take weeks or months. Urgent decisions take hours.
  • The cost of delay. Replying to an elective lead in four hours instead of one is a missed marketing opportunity. Replying slowly to an urgent one can be a clinical problem.
  • Who should respond. Elective leads suit a counsellor. Urgent ones need a clinician-approved triage path.

Step 1: Classify before you score

Add a routing step that assigns each enquiry to a class. Scoring happens inside the class, never across classes.

ClassExample signalsHandled by
Possible urgentSymptoms described as sudden or severe, chest pain, breathlessness, inability to bear weight after a fall, "emergency", "tonight"Clinician-approved triage script; human on duty
Time-sensitive electiveKnown diagnosis, surgeon recommended a date, insurance pre-authorisation pending, travelling soonSenior counsellor
High-consideration electiveResearch phase, comparing surgeons, cost and recovery questionsCounsellor with nurture sequence
Information onlyCareers, billing queries, reports, wrong numberRelevant desk

The signals list for "possible urgent" must be written and signed off by your clinical leads, not by marketing or an algorithm. Make it conservative. A false alarm costs a short call. A missed one can cost far more. For cardiology in particular, the safe public message for symptoms such as chest pain is to seek emergency care immediately rather than wait for a clinic reply, and your forms, auto-replies and chatbot should say so in clinician-approved words. In India the unified emergency number is 112.

NMC's October 2026 advertising guidelines also warn against promotion that induces unnecessary investigations or procedures through discounts or inducements. That bears on scoring too. A lead score should rank by likely fit and timing, never by how much a patient can be nudged into an unneeded procedure. Ask your compliance lead to review what the score is allowed to trigger.

Step 2: Score the elective classes separately

High-consideration surgery (joint replacement, spine, bariatric)

Here, intent builds up across weeks, so a behavioural score is meaningful. A sample rule set, with invented weights to show the structure:

SignalPoints
Asked about a specific procedure by name+15
Viewed surgeon profile page+10
Viewed cost or insurance page+15
Uploaded or mentioned a recent scan+20
Returned within seven days+10
Asked about timing ("when can I be operated")+15
Located within travel distance+10
Generic "price list" request, no clinical context+0
Only downloaded a general guide+5
No reply to two follow-ups-15

Cap the score at 100 and band it: 70 or more is "ready", 40 to 69 is "warming", under 40 is "nurture". These weights are placeholders. The next step in a real project is to fit or adjust them against your own arrivals data, and to check them as described in our post on validating a lead score.

Time-sensitive elective

These are patients with a recommendation and a deadline. Score on practical blockers rather than interest: insurance status, referral present, surgeon availability, travel dates. The question is not "how interested?" but "what stands between this person and a date?" A checklist of blockers with an owner for each beats a numeric score here.

Step 3: Treat urgent as a pathway, not a score

Do not rank urgent enquiries against others. Route them. The logic is a rule, not a model:

  1. Any enquiry flagged as possibly urgent bypasses the queue.
  2. The first response is the clinician-approved message advising emergency care where appropriate, delivered immediately, even outside working hours.
  3. A human on duty sees the item within a defined time, and the CRM records who acted and when.
  4. Nothing about the enquiry enters a marketing nurture flow.

Marketing automation should never message a possible-urgent contact with promotional content. Suppress them from campaigns until a clinician or senior staff member clears the flag.

Want this applied to your practice? Get a free ai & automation audit.

Book my free audit

Step 4: Set response targets by class

Service-level targets belong to the class, and operations leadership should set them. Example targets, to adapt:

ClassFirst response targetOwner
Possible urgentImmediate auto-reply with approved safety message; human review within minutes during staffed hoursDuty nurse or coordinator
Time-sensitive electiveSame working day, under two hours in office timeSenior counsellor
High-consideration elective, ready bandWithin one working hourCounsellor
High-consideration elective, warmingWithin one working dayCounsellor
NurtureAutomated sequence, human reply on engagementMarketing

Measure breach rates per class, not an average. An average first-response time of thirty minutes can hide the fact that every night-time urgent message waits until morning. Where your own CRM makes a response clock visible per class, Pioneer CRM is one place to keep enquiries, owners and follow-ups together; the design works with any tool that can hold a class field and a timestamp.

Step 5: Build the data model

Add these fields so the classes are reportable:

  • enquiry_class: possibleurgent, timesensitive, highconsideration, infoonly
  • class_source: rule, counsellor or clinician (who decided)
  • class_changed_at and class_changed_by
  • score and score_version, for elective classes only
  • first_response_at and first_response_by
  • clinician_reviewed: boolean, mandatory for possible_urgent
  • outcome: booked, attended, procedure done, lost, referred out

The class field lets you answer questions a single score cannot, such as how many elective leads were wrongly flagged urgent, and how many urgent ones were first noticed after hours.

Failure modes

  • Class creep. Counsellors tag everything "urgent" to jump the queue. Audit the flag monthly and compare against clinician review.
  • Keyword brittleness. Patients describe symptoms in Hindi, Hinglish and their own words. Test the routing rules on real, anonymised messages in each language.
  • Channel gaps. WhatsApp, calls and web forms may each have a different path. The urgent pathway must work on all.
  • Privacy. Symptom text is sensitive personal data. Limit access, set retention rules and take advice on the DPDP Act obligations that apply to your practice.
  • Training on the wrong label. If you later build a model, train it only on elective classes. Including urgent items teaches it that fast, plain messages are low value.

Where to start

  1. Ask clinical leads to write the possible-urgent signals and the approved first message.
  2. Add the enquiry_class field to your CRM and tag the last 200 enquiries by hand.
  3. Count how the classes split. Count them before designing anything; the cost of getting this class wrong is higher than its share of volume suggests.
  4. Create separate scoring for high-consideration elective leads, with invented weights replaced by your data over time.
  5. Set per-class response targets and report breaches weekly.

Our pages on AI lead scoring, lead pipeline management and the lead scoring glossary entry cover the basics. For specialty context, see the marketing notes for orthopaedic joint-replacement practices and cardiology.

Filed underlead scoring elective surgeryurgent vs elective patient leadsjoint replacement lead scoringhospital enquiry triage CRMlead response time by class

About the author

Co-Founder & CTO · Gurugram, India

Arush builds the systems the rest of the company runs on — the patient-acquisition pipelines, the AI chatbots, the analytics that tie a single click to a booked appointment. If a campaign reports a number, it's because something his team built is tracking it.

Free guide · 15 pages

The Patient Acquisition Blueprint

A 90-day plan from first search to booked appointment.

The Pioneers Brief — one essay, three benchmarks, one teardown, every Thursday. Subscribe free →

Free ai & automation audit

Your next chapter starts with a conversation.

Three quick questions. We come prepared with a plan for your specialty and city.

  • A 30-minute call with a healthcare specialist
  • Your top three growth opportunities, in writing
  • Honest advice — even if that means not hiring us
  • No obligation and no hard sell
You’ll speak with a senior strategist
Consultations are hosted by Arush Thapar’s team.
Step 1 of 3Takes 30 seconds

What do you want to improve first?

Details stay privateNo obligationReply within 1 working day

Client portfolio · Healthcare websites

Work relevant to scoring elective vs urgent leads why one score fails joint.

Explore client work →