Adult cow mortality

Definition

Adult cow mortality is the rate at which cows that have calved at least once die on the farm, expressed as an annualized percentage of the average herd. Deaths include cows found dead and cows euthanized on the farm. Cows sold, even if they are sold because they are sick and slaughtered soon after, are removals, not deaths (see cull rate).

The rate is per cow-year, not a share of individual cows: with turnover, more cows than the average herd size pass through in a year.

Cause coding

Each death should carry a cause, entered with the death event. A workable short list:

Calving or fresh-cow related (dystocia, milk fever, metritis, DA, ketosis)
Mastitis
Lameness or injury (including downer cows)
Digestive (other than DA), respiratory
Other, and unknown

Report the share coded “unknown”; above about a quarter, the cause breakdown is not reliable. Also report deaths by DIM: deaths in the first 60 DIM are a transition measure.

Formula

Plain text:

Cow-days method:
  Annual death rate (%) = deaths in period / (sum of daily adult cow counts in period / 365) × 100

Monthly × 12 method:
  Annual death rate (%) = deaths in month / average adult cows in month × 12 × 100

LaTeX:

\mathrm{DR}_{\text{annual}} = \frac{D}{\sum_{d} N_{d} / 365} \times 100
\qquad
\mathrm{DR}_{\text{month}\times 12} = \frac{D_{\text{month}}}{\bar{N}_{\text{month}}} \times 12 \times 100

The cow-days method is the more accurate one: it uses the actual time cows were at risk and works for any period length. The monthly × 12 method treats all months as equal (a 28-day month and a 31-day month count the same) and is very noisy for one month in a small herd. Both give nearly the same answer over a full year with a steady herd size. State which method and which cow count (lactating only, or lactating plus dry) is used; see cow count definitions.

Example: 6 deaths in a 30-day month with 1,200 cows on hand every day gives 6 / (36,000 / 365) × 100 = 6.1% annualized. The monthly × 12 method gives 6 / 1,200 × 12 × 100 = 6.0%.

How to pull it in DC305

Deaths are recorded with a DIED (or DEAD) event, usually with a remark code for the cause; euthanasia may be a separate remark or event. Some herds record deaths as a sold event with a “died” code. Check the herd’s private dc305-vocab mapping for the event and cause codes.

EVENTS\D DIED ID LACT DIM FOR LACT>0     DIED events, with remark (cause); animals that left the herd only
LIST\D ID LACT DIM DIEDT FOR LACT>0 DIEDT>-365   cows that died in the last year, with died date (DIEDT)
ECON\E                                   dead cow summary report
COUNT FOR LACT>0                         adult cows on hand

EVENTS lists live animals only by default, so a list of deaths without \D (dead only) or \B (live and dead) comes back empty or short. \S prompts for a date range (\S### starts ### days ago); event tables default to the last 365 days.
Animals moved to the archive file are read with \L.
For the denominator, use daily inventory if available (ECON\ID gives a past or yearly cow inventory summary, e.g. ECON\ID750 for the past 750 days); otherwise the average of start and end counts for each month, summed as cow-days.
Cross-tab the cause remark and DIM at death.

Typical range

Herd context: US Holsteins, freestall or dry-lot, all adult cows (lactating and dry).

Measure	Typical	Goal
Annual adult cow death rate	5–9%	< 5%
Share of deaths in first 60 DIM	25–40%	lower is better

US reference (NAHMS, operations with 30+ cows, deaths as a percentage of January 1 cow inventory): 3.8% in 1996, 4.8% in 2002, 5.7% in 2007 and 5.6% in 2014 (NAHMS Dairy 2014 trends report, table F.3.b).

Herd size has a small and inconsistent effect. In NAHMS Dairy 2007, large herds (500+ cows) reported more deaths than small herds (6.1% vs 4.9%); in Dairy 2014 the gap (5.8% vs 5.0%) was not significant. On-farm euthanasia rather than selling compromised cows, and recording differences, both raise recorded deaths (Thomsen & Houe 2006). McConnel et al. 2008 linked higher herd mortality to more lameness, respiratory disease and antibiotic treatment, TMR feeding, fewer culls before 50 DIM, longer calving intervals and region.

Common mistakes

Mixing the two methods. Monthly × 12 for one month and cow-days for a year won’t tie exactly; label the method.
Wrong denominator. Lactating cows only vs lactating plus dry changes the rate by roughly 10–15%. Heifers are never in the denominator.
Counting sold-to-slaughter as dead, or dead as sold. Shifts between euthanasia and selling move the death rate without any change in cow health. Look at death rate and cull rate together.
Missing deaths. Deaths recorded only in a separate log or entered late never reach the event list; check the count against rendering pickups.
Annualizing one month in a small herd. Two deaths in a 200-cow herd in one month annualizes to 12%. Use rolling 12-month figures.
Treating “unknown” as a cause. It is missing data.