Study guide

10.4 Calculating Injury Incidence & Frequency Rates

IOSH Managing Safely Exam

Track your progress as you read

Finished sections get a check mark.

1 Introduction & Exam Overview

1.1 What Is IOSH Managing Safely? 1.2 Exam Format & Dual Assessment 1.3 Certificate Validity & Study Plan

2 Chapter 1: Introducing Managing Safely

2.1 Moral, Legal & Financial Reasons 2.2 The Accident Cost Iceberg 2.3 Responsibility & Accountability

3 Chapter 2: Assessing Risks

3.1 Hazard, Hazardous Event & Risk 3.2 Five Steps of Risk Assessment 3.3 The 5×5 Likelihood × Consequence Matrix 3.4 Residual Risk, Suitable & Sufficient, Review

4 Chapter 3: Controlling Risks

4.1 ERICPD Hierarchy of Control 4.2 Reasonably Practicable 4.3 Safe Systems of Work (SSoW) 4.4 Permits to Work

5 Chapter 4: Understanding Responsibilities

5.1 Criminal vs Civil Law 5.2 Employer & Employee Duties 5.3 Enforcement, Improvement & Prohibition Notices 5.4 Knowledge Tests & Management Systems (HSG65 / ISO 45001)

6 Chapter 5: Hazard Categories, Noise, Vibration & Substances

6.1 Six Hazard Categories 6.2 Noise Hazards & Action Values 6.3 Hand-Arm & Whole-Body Vibration 6.4 Hazardous Substances, Asbestos & LEV

7 Chapter 6: Work at Height, Manual Handling, Fire & Electricity

7.1 Work at Height 7.2 Manual Handling, Slips & Trips 7.3 Fire Safety Basics 7.4 Electricity Hazards

8 Chapter 7: Stress, Work Equipment & Other Hazards

8.1 Work-Related Stress & Psychosocial Hazards 8.2 Work Equipment & Machinery 8.3 Confined Spaces, Workplace Transport & Welfare

9 Chapter 8: Investigating Incidents

9.1 Accidents, Incidents & Near Misses 9.2 Immediate, Underlying & Root Causes 9.3 Four-Step Investigation Process 9.4 RIDDOR Reporting Overview

10 Chapter 9: Measuring Performance

10.1 Active (Proactive) vs Reactive Indicators 10.2 KPIs & Characteristics of Good Metrics 10.3 Auditing the Safety Management System 10.4 Calculating Injury Incidence & Frequency Rates 10.5 Inspections & Benchmarking

11 Chapter 10: Protecting the Environment

11.1 Environmental Basics: Pollution, Waste & Duty of Care 11.2 Reducing Environmental Impact & Environmental Management Systems

12 Chapter 11: Risk Assessment Project & Exam Day

12.1 Project Requirements & Marking 12.2 Worked Example: Four-Hazard Assessment 12.3 Written Assessment Strategy & High-Yield Review

10.4 Calculating Injury Incidence & Frequency Rates

Key Takeaways Explain This Topic AI

Injury incidence rate = (number of reportable injuries in the period ÷ average number of employees in that period) × 100,000.

Worked example: 2 reportable injuries among an average of 50 employees gives (2 ÷ 50) × 100,000 = 4,000 per 100,000.

Rates normalise for size, so a small site with few injuries can have a far worse rate than a large site with more injuries.

Incidence rates divide by people employed; accident and lost-time frequency rates divide by hours worked, so the two are not interchangeable.

Rates are lagging, swing wildly on small numbers, are flattered by under-reporting, and never explain cause — pair them with active measures.

Last updated: August 2026

10.4 Calculating Injury Incidence & Frequency Rates

Quick Answer: Raw accident counts mislead because they ignore how many people are exposed. Managing Safely teaches you to convert reactive data into a rate . The injury incidence rate is (number of reportable injuries in the period ÷ average number of employees in that period) × 100,000 . Two reportable injuries among 50 employees is 4,000 per 100,000 — far above a typical national rate. Frequency rates normalise by hours worked instead of headcount. Rates let you benchmark; they still cannot tell you why .

Section 10.1 separated active from reactive measures. This section takes the reactive half and makes it usable. Managing Safely does not stop at "count your accidents" — the module demonstrates how to calculate a rate so one site can be compared with another, and with the industry. It is one of the few genuine calculations on the course, which makes it a reliable exam item and a skill most line managers never acquire.

Why a raw count is not a performance measure

Consider two depots reporting to the same operations director:

  • Depot Reportable injuries last year People employed
  • Northgate 6 900
  • Southgate 3 40

Northgate reports twice as many injuries, so the naive reading is that Northgate is the problem site. It is the wrong reading. Southgate injured 3 of 40 people ; Northgate injured 6 of 900 . Until you divide by exposure, you are measuring size , not safety . Every large site looks bad and every small site looks good, and managers of small high-risk teams escape scrutiny.

Normalising also protects you in the other direction. A site that grows from 60 to 300 staff will see its accident count rise even if its control has improved. Without a rate, growth looks like decline.

The injury incidence rate (the Managing Safely calculation)

The incidence rate normalises by people :

Injury incidence rate = (number of reportable injuries in a given time period ÷ average number of employees during that period) × 100,000

Three things to hold onto:

The period is normally one year , so numerator and denominator describe the same window.

Average number of employees — not the headcount on the day you ran the report. Sites with seasonal peaks must average honestly.

The × 100,000 is only a scaling base. It converts a tiny decimal into a number people can talk about, and it matches the per-100,000 basis on which HSE publishes national and industry injury statistics — which is exactly what makes your figure comparable.

  • Worked calculation 1 — small workshop
  • 2 reportable injuries in the year
  • Average 50 employees
  • (2 ÷ 50) × 100,000 = 0.04 × 100,000 = 4,000 per 100,000 employees
  • Worked calculation 2 — the two depots
  • Depot Injuries Average employees Calculation Incidence rate
  • Northgate 6 900 (6 ÷ 900) × 100,000 667
  • Southgate 3 40 (3 ÷ 40) × 100,000 7,500

Southgate's rate is over eleven times Northgate's. The count said Northgate; the rate says Southgate. That reversal is the whole point of the calculation, and it is exactly the shape of scenario an exam item uses.

Worked calculation 3 — comparing with the industry

Your company: 5 reportable injuries, average 400 employees.

(5 ÷ 400) × 100,000 = 1,250 per 100,000

If the published rate for your industry is around 500 per 100,000, you are roughly two and a half times the sector figure. That is the sentence to take to a management meeting — not "we had five accidents".

Incidence rate vs frequency rate — the exam discriminator

Both normalise; they differ in what they divide by .

Measure Denominator Formula Answers

Incidence rate People employed (injuries ÷ average employees) × 100,000 How many of our people got hurt?

Accident frequency rate (AFR) Hours worked (accidents ÷ total hours worked) × 100,000 How many injuries per unit of work done ?

Lost-time injury frequency rate (LTIFR) Hours worked (lost-time injuries ÷ total hours worked) × 1,000,000 How often does work stop people working?

Severity measure Days lost days lost ÷ number of injuries (or per hours worked) How serious were the outcomes?

Practical differences that matter to a manager:

Hours-based rates handle part-time, shift and overtime patterns better. A team working heavy overtime is exposed for longer than its headcount suggests, and only an hours-based rate shows that.

People-based rates are easier to produce — payroll gives you the denominator without a timesheet system, which is why the incidence rate is the one taught on the course.

The multiplier is part of the definition. UK frequency rates commonly use 100,000 or 1,000,000 hours; the US OSHA convention uses 200,000 hours. A rate calculated on one base cannot be compared with a rate calculated on another.

Using rates without being fooled by them

Rates are lagging measures. They are useful, and they are limited:

Limitation What goes wrong Manager response

Small numbers swing wildly One injury in a 25-person team moves the rate by 4,000 Look at trends over several years; never manage a small site on a single year's rate

Under-reporting flatters the number Pressure to keep the rate down suppresses reports Watch near-miss volume alongside; a falling injury rate with falling near-miss reports is a warning, not a win

Silent on ill health Noise, vibration, stress and substance harm surface years later and rarely appear as injuries Track health surveillance findings and absence data separately

Silent on cause The rate never says why Pair every rate with investigation findings and active measures

Comparability traps Different bases, different definitions of "reportable" State the formula and the base every time you publish a figure

A depot with a zero incidence rate and no inspections, no in-date risk assessments and no near-miss reports has not proved it is safe. It has proved that nobody has been hurt yet .

Turning a rate into management action

Calculate consistently — same formula, same base, same definition of reportable, every period.

Trend it — three to five years beats one snapshot.

Benchmark it — against your other sites and the published industry figure (Section 10.5).

Explain movement — did the rate fall because control improved, or because a contract ended and exposure fell?

Pair it with leading indicators — inspection completion, risk-assessment currency, action closure (Section 10.2).

Never set it as a bonus target on its own — that is the fastest route to under-reporting.

Common exam traps

  • Trap Correction
  • Dividing employees by injuries Injuries go on top ; employees underneath
  • Forgetting the × 100,000 Without the base you get 0.04, not 4,000
  • Using today's headcount for a full year The formula asks for the average number employed over the period

Comparing an incidence rate with a frequency rate Different denominators — people vs hours; the numbers are not interchangeable

  • Mixing a per-100,000 rate with a per-200,000 or per-1,000,000 rate State and match the base before comparing
  • Treating a low rate as proof of good management Low rates can reflect luck, small size or suppressed reporting
  • Manager checklist

Can I write the incidence rate formula from memory, including the × 100,000 ?

Do I use the average number employed, not a convenient snapshot?

Do I calculate the same way every period so the trend means something?

Do I compare against a published figure on the same basis?

Do I read the rate next to near-miss volume, health surveillance and active measures?

Do I resist targets and bonuses that reward a low number rather than good control?

Get comfortable with this one calculation and the whole Measuring Performance module becomes concrete: reactive data in, comparable number out, management conversation improved.