Bradford Factor Calculator: Score & Thresholds

The Bradford Factor is an attendance-management score designed to highlight frequent short sickness absences. It assumes that several separate disruptions can be harder to cover than one absence of the same total length. Employers sometimes use it as a consistent prompt for a return-to-work conversation or attendance review.

It has no status in UK employment law and there is no official “good” or “bad” score. The organisation chooses the measurement period, what counts as an episode and any review points. Those choices should be documented, communicated and applied consistently, with reasonable adjustments and protected absences considered separately.

Enter each separate continuous spell as one episode. If the included episode lengths are 1, 2 and 4 working days, S is 3 and D is 7. The score is 3² × 7 = 63. The list makes the arithmetic auditable because a manager can see which spells produced the count and total.

Overlapping records should not be double-counted. A weekend or non-working day inside one continuing sickness spell does not automatically create a new episode. If the employee returned to work and was later absent again, that will normally be a new episode, subject to the employer's written policy.

Squaring S makes the score sensitive to frequency. The same ten lost working days can produce scores from 10 to 1,000 depending on how they are grouped.

This weighting is both the purpose and the limitation of the formula. It describes a pattern; it does not explain the cause, predict future attendance or prove misconduct.

No government body sets Bradford Factor thresholds. The bands below are illustrative, non-statutory examples, not recommended sanctions. An employer should choose review points that fit its operation, workforce and equality duties, then state them in the absence policy.

These labels mirror the calculator's example bands, not a lawful automatic sequence. Crossing a band should open a fair review, not determine its outcome. Length of service, cause, warnings, medical evidence and previous support remain relevant.

The Bradford Factor measures frequency and duration; it does not calculate the employer's financial cost. Two patterns can have the same total days absent and therefore a similar direct sick-pay cost, while producing very different scores because S is squared. For example, one five-day episode scores 5, while five one-day episodes score 125.

For an absence-cost estimate, keep the employee's normal weekly earnings, qualifying days, enhanced company sick pay and any cover or overtime costs separate from the Bradford score. The [SSP calculator](/ssp-calculator/) can model statutory sick pay, while the [company sick pay calculator](/sick-pay-calculator/) compares SSP with enhanced policies. Do not infer a pound cost from a threshold band alone.

The policy should define whether D counts scheduled working days, calendar days, shifts or part-days. Most implementations use working time actually missed, excluding rest days that the employee was not due to work. A four-day sickness spanning a weekend may therefore contain two lost working days but remain one continuous episode.

Planned annual leave, authorised family leave, jury service and other non-sickness leave should not be silently mixed into the score. A fit note supports sickness absence but does not by itself decide whether an episode is included. Keep the raw absence record separate so adjustments do not erase the underlying facts.

Acas says employers should consider recording pregnancy-related and disability-related absence separately. Under the Equality Act 2010, reasonable adjustments for a disabled employee can include ignoring some disability-related absence, increasing a review trigger or using a different process. The correct adjustment is individual, not a blanket rule.

Pregnancy-related sickness must not lead to unlawful unfavourable treatment. Long-term conditions, treatment schedules and mental health episodes can also create high frequency scores without misconduct. Occupational health advice may help, but consent, confidentiality and the employee's explanation still matter. Never assume a high score removes the need to investigate and support.

A defensible process starts with a written policy, accurate records and the same calculation window for comparable employees. Managers should verify dates with the employee, hold a return-to-work or review meeting, explore causes and support, and record why any action is proportionate. Acas describes trigger points as prompts for review when used appropriately.

Sickness details can reveal health information, which requires careful access controls and retention under data-protection rules. Share the score only with people who need it. Audit changes to episodes and exclusions so the result can be explained, corrected and distinguished from the separate managerial decision.

Choose the policy period, add one row per separate sickness episode and check the working days lost. The calculator counts episodes as S, totals days as D and returns S² × D. If an episode is duplicated or one continuous spell is split, the squared frequency can distort the result sharply.

Compare the score only with the employer's published non-statutory bands. Before any decision, check protected and disability-related absence, reasonable adjustments, record accuracy and the employee's explanation. The calculator measures frequency and duration; it does not decide fairness, capability or conduct.