Absences by Duration Group

Short-Term or Long-Term Illness: How Do I Correctly Analyze Absences by Duration Group?

Under German law, sick-leave cases are grouped by duration: 1–3, 4–7, 8–14, 15–41, and 42+ days. Many short cases and a few long ones can produce the same sickness rate but have different causes. Continued pay ends after six weeks of incapacity to work; anyone who is unable to work for more than six weeks within one year, continuously or repeatedly, must be offered BEM by the employer.

The Five Duration Groups and What They Mean

Group	Duration (Days)	Typical Pattern	Legal Threshold
Short-term	1–3	Frequent short illnesses; if recurring, an early indicator of strain or motivation issues	Sick note generally required only from day 4 (Sec. 5 EntgFG)
Short-Medium	4–7	Typical acute illnesses lasting up to about a week	Physician's sick note is required
Medium	8–14	Longer acute episodes lasting up to about two weeks	—
Long	15–41	Extended illness still within the continued-pay period	Still within the 6-week continued-pay period (Sec. 3 EntgFG)
Long-term	42+	Long-term illness; often chronic or mental-health conditions	Continued pay ends, sick pay begins (Sec. 44 ff. SGB V); BEM entitlement (Sec. 167 (2) SGB IX)

Why doesn't the sickness rate alone explain anything?

Two companies with 100 employees each and an identical 5 percent sickness rate can have completely different problems: In the first, hundreds of one- to three-day cases pile up — a pattern that can point to workload, leadership climate, or simply waves of infection. In the second, fifteen long-term cases account for almost all the lost time — here the issue is chronic illness, return-to-work management, and BEM.

The rate is the temperature reading; duration-group analysis is the diagnosis: for each sick-leave case, the duration is calculated (end date minus start date) and the case is assigned to one of five classes. Two distributions are then evaluated — number of cases per group and number of days per group. Together they form the picture: short cases almost always dominate the case count, long-term cases almost always dominate the days.

This requires individual sick-leave cases with a start and end date — a mere annual total of sick days is not sufficient for this analysis. Anyone with only totals from payroll should plan individual-case tracking as the first step in expanding their reporting.

What happens after six weeks of incapacity for work?

After six weeks of continuous incapacity to work, the employer's continued pay obligation ends (Sec. 3 EntgFG, German Continued Pay Act). From that point, the statutory health insurer pays sick pay — typically 70 percent of gross pay, capped at 90 percent of net pay (Sec. 47 SGB V). For the company's cost accounting, day 42 is a genuine turning point; for the affected employee, it means a noticeable drop in income.

At the same time, the BEM duty arises: anyone who has been unable to work for more than six weeks, whether continuously or repeatedly, within one year must be offered a Betriebliches Eingliederungsmanagement (company reintegration management) by the employer (Sec. 167 (2) SGB IX). Important: this threshold also applies to repeated short illnesses that add up over twelve months — not only to one continuous long-term case. That is exactly why determining BEM eligibility requires the same individual sick-leave case data as duration-group analysis.

For reporting purposes, it is therefore advisable always to report the “42+” group separately: it is simultaneously a cost driver, the transition to sick pay, and a BEM trigger — the one figure that matters equally to management, HR, and the works council.

How do you read short-term patterns without putting employees under general suspicion?

A noticeable cluster of one- to three-day cases is a signal, not proof: it can point to overload, team conflict, or leadership problems — but just as easily to a flu wave, daycare closures, or simply seasonal effects. Serious analysis therefore always compares over time (same period, prior year) and across comparable units before drawing conclusions.

The Bradford Factor — frequency squared × total days — is a widely used tool for weighting frequent short absences more heavily than rare long ones. It works as a screening tool at the level of organizational units; as an assessment tool for individuals it is contentious and, under German employment law, should be used with caution. Recommendation: use it in aggregate as an early-warning signal, never communicate it as an individual metric.

On the intervention side, short-term patterns belong in condition-based prevention: psychosocial risk assessment, leadership-awareness training, work organization. Anyone who instead responds with attendance bonuses or pressure around returning from illness shifts absences into presenteeism — working while sick, whose costs the German Federal Institute for Occupational Safety and Health (BAuA) ranks as at least on par with absenteeism.

What helps when long-term cases dominate the days lost?

When groups 15–41 and 42+ dominate the lost days, the lever shifts from broad programs to case-by-case work: consistently offering BEM to everyone entitled to it, phased return to work under the Hamburg Model, and staying in touch during the illness — voluntary, caring, without pressure.

A structural driver of long case durations is mental illness: TK (Techniker Krankenkasse) reports a long-term rise in absences due to psychological disorders of 159 percent since 2006 (age-adjusted, TK Health Report 2025). Mental-health diagnoses lead to above-average case durations — anyone seeing many 42+ cases should prioritize psychosocial risk assessment and corresponding prevention programs.

Important for interpretation: a few long-term cases can statistically dominate a small workforce. A single twelve-month case corresponds to around five percentage points of sickness rate in a 20-person team (1 in 20 people absent all year) — so never judge trends in small units on a single year.

Is duration-group analysis permissible under data protection law?

Yes — duration-group analysis is permissible because, and as long as, it stays aggregated: cases and days per group, per reporting unit — never names, never diagnoses, never individual case histories in reports. Sick-leave data is health data under Art. 9 GDPR.

We recommend a minimum group size of five people (a common convention, not a statutory figure): a reporting cell (e.g., “Department X, Group 42+”) is suppressed if fewer than five employees are behind it — otherwise a specific person's long-term illness becomes effectively identifiable. This is quickly the case especially with long-term cases in small departments; reporting then drops to a coarser level (site instead of department).

Identifying BEM-eligible individuals by name must be treated separately: it is permissible as a statutory duty, but belongs in an access-restricted HR process — not in KPI reports.

Related measures & topics

Key takeaways

Same sickness rate, different diagnosis: only duration groups (1–3 / 4–7 / 8–14 / 15–41 / 42+) show what is actually driving your absences.
Always evaluate both: cases per group (short cases dominate) and days per group (long-term cases dominate).
Report 42+ separately: end of continued pay (Sec. 3 EntgFG), sick pay (Sec. 47 SGB V), and BEM entitlement (Sec. 167 SGB IX) all coincide.
The BEM threshold also applies to short sick-leave periods that add up within 12 months — not only to continuous long-term cases.
Short-term patterns → condition-based prevention and psychosocial risk assessment; long-term patterns → BEM and phased return to work.
Mental illness drives long case durations: +159% in absences since 2006 (TK, age-adjusted).
GDPR: only evaluate in aggregate, suppress cells under 5 people — especially for 42+ in small units.

Frequently asked questions

From when is a sick note required?+

Statutory standard: if incapacity to work lasts longer than three calendar days, a physician's certificate must be provided by the following working day at the latest (Sec. 5 EntgFG). The employer can also require it earlier. As a result, one- to three-day cases are often undercounted in the data if only certified sick-leave cases are counted — take this into account when interpreting the short-term group.

For the six-week BEM threshold, do calendar days or working days count?+

Sec. 167 (2) SGB IX refers to being “unable to work for more than six weeks, continuously or repeatedly” within one year and names no number of days. Whether calendar days or working days count has not been settled by the highest courts. For a single continuous case the difference is small; with many repeated short periods of incapacity it can decide whether the threshold has already been reached. Recommendation: count generously (when in doubt, offer BEM) — offering BEM too early is not a risk, failing to offer it is.

What is the Bradford Factor and should we use it?+

Bradford Factor = number of sick-leave episodes squared × total days. It heavily overweights frequent short absences and makes short-term patterns visible that get lost in plain day totals. Useful as an aggregated early-warning signal per organizational unit; unsuitable and legally risky as a basis for individual assessment or sanctions.

Our short-term cases are clustering in one department — what should we do?+

Validate first, then act: check the prior-year comparison and seasonal effects, and observe the minimum group size. If the pattern is confirmed, the next step is not individual-case monitoring but a psychosocial risk assessment for that area, plus a structured conversation with the manager about work organization and climate.

Are we allowed to contact employees while they are on sick leave?+

In principle yes — as a gesture of care, not control: no questions about diagnosis, no requests to work, no recovery deadlines. When done right, staying in touch is experienced by many employees as appreciative and eases their return. Standardized return-to-work conversations may be subject to works council co-determination — details and sample wording are in the BEM-Kompass guide “Contact During Sick Leave.”

Automatically Evaluate Duration Groups

EasyWHP automatically classifies sick-leave cases into duration groups, flags BEM candidates early as a precaution (from 30 sick days in 12 months, with weekends within a sick note counted), and maintains the minimum group size — without manual Excel work.

Start for Free Check WHP Maturity Level

Sources

Last updated: 2026-07-07. Not legal or tax advice — have your specific case reviewed by a professional.

WHP KPIs: The Basic SetWhich KPIs belong in a basic workplace health promotion (WHP) reporting set — and how are they calculated consistently?BEM deadline: six weeksBEM after more than six weeks of incapacity for work: When must the employer offer the meeting?Psych Risk Assessment for SMEsDoes a small or mid-sized business (SME) have to conduct a risk assessment for psychological workload?TK Health Report 2025What does the TK Health Report 2025 say — and what does it mean for employers?BAuA: Presenteeism CostsWhat are presenteeism costs — and why can they exceed absenteeism costs according to BAuA?