Abstract

The purpose of this study was to determine clinical disease incidence in eight non-seasonally

calving, pasture-fed dairy herds in the southern highlands district of New South Wales. This was a

longitudinal population study. The study included all cows that calved between January 1994 and

December 1995 and consisted of 2111 lactation records from 1430 cows. The incidence of the more

common diseases were: calving-associated disorders, 18.0 cases per 100 calvings (95% CI 16.4±

19.8 cases per 100 calvings),  metabolic disorders, 5.5 cases per 100 cow-yr at risk (95% CI 4.5±6.6

cases per 100 cow-yr at risk),  reproductive-tract disorders, 22.3 cases per 100 cow-yr at risk (95%

CI 19.2±25.8 cases per 100 cow-yr at risk),  udder disorders, 17.6 cases per 100 cow-yr at risk (95%

CI 15.9±19.5 cow-yr at risk) and lameness, 3.7 cases per 100 cow-yr at risk (95% CI 2.9±4.7 cow-yr

at risk). In agreement with dairy-cow disease-incidence studies conducted elsewhere, disorders of

the reproductive-tract and udder were the most frequent clinical conditions encountered. These

findings confirm that dairy herd-health programs should emphasise the control of these two groups

of disorders. # 2000 Elsevier Science B.V. All rights reserved.

1. Introduction

A benefit of monitoring disease events as part of a herd-health package offered to dairy

farmers is that valuable information is accumulated on all diseases that are recognised by

either the farmer or the veterinarian. This process provides information of the incidence

of clinical disease in an environment where accurate and complete recording of disease

events can be anticipated (Williamson et al., 1978). A knowledge of the expected levels

of clinical disease that should be encountered in dairy herds allows veterinarians

confidently to identify situations where higher-than-expected incidences of disease exist

and assists in determining priorities for preventive programs. Relating the disease

incidence to parity, stage of lactation, previous disease history, and level of production

identifies those individuals at risk of developing disease conditions and this assists in the

further refinement of herd-health care programs.

In spite of the increased use of microcomputers to record cow event and health data

since the mid-1980s (a situation that should facilitate the recording and reporting of

dairy-cow disease patterns), studies of disease incidence in Australian dairy herds have

been few and published studies have been restricted to those conducted in the late 1960s

and 1970s (McClure and Dowell, 1968,  Williamson et al., 1978).

The purpose of this study was to determine clinical disease incidence in dairy herds

receiving programmed herd-health visits in the southern highlands district of New South

Wales, Australia. While purposive selection of herds in this study resulted in a group of

higher-producing herds compared to the state average, the data in this study nonetheless

provide useful estimates of dairy-cow disease incidence under New South Wales farming

conditions. With current interest in food safety, it is conceivable that in the near future,

producers of milk for human consumption will be required to provide transparent

documentation of herd disease status. A secondary objective of this study was to evaluate

the suitability of the disease classification system used in this study for this purpose.

2. Materials and methods

This study was conducted on eight non-seasonally calving dairy herds located in the

author’s area of practice, in the southern highlands district of New South Wales, Australia

(Fig. 1).

This was a prospective longitudinal population study. Herd selection was purposive and

was based on the herd manager’s demonstrated ability to maintain accurate herd records.

During 1992 and 1993, each participant herd began a program of monthly health visits

under the direction of the author. The primary purpose of these visits were to maintain

herd reproductive efficiency by pregnancy testing cows thought to be in calf and by

examining and treating reproductive disorders (Blood et al., 1978,  Esslemont, 1994). In

some herds, advice was provided on nutritional management, the maintenance of milk

quality and the management of replacement stock.

As part of the herd-health program provided, the biographical details of all stock on the

farm were entered into the computerised database DairyCHAMP (Udomprasert and

Williamson, 1990) and managers were requested to record all important production and

health events into a dairy (Williamson et al., 1978). Data entered into the dairies for each

individual included dates and descriptions of calvings, heats, services, disease events,

treatment events, dry-off events, culling events and the associated reasons for culling.

Milk-production records from the New South Wales Agriculture Dairy Herd Improvement

(DHI) scheme for each participant herd were downloaded directly into the herd

database. Milk-production recording was conducted monthly and included (for each

2 M.A. Stevenson / Preventive Veterinary Medicine 43 (2000) 1±11

lactating cow) test-day milk yield (l), test-day fat yield (kg), and test-day protein yield

(kg).

Prior to each scheduled monthly herd-health visit, information from the herd dairies

was entered into the database by the author. At the time of data entry, a check was made

that event information provided was consistent with each cow’s previous event history.

Where discrepancies were found, clarification was made with manager during the herdhealth

visit and the corrected event details were re-entered into the database.

Herd managers were instructed to record only disease events that required treatment or

direct intervention, according to the definition schedule provided in Table 1. Where a

veterinarian attended an animal after a diagnosis was initially recorded by the manager,

the diagnosis provided by the veterinarian replaced the initial diagnosis, where

appropriate. Managers varied in their ability to differentiate between the different

conditions associated with lameness and for this reason lameness disorders were

classified according to the location of the causative lesion (that is either the foot, the leg,

or elsewhere). The term cystic ovarian disease was used to identify the situation where a

cow showed either irregularly short or long interoestrus intervals along with abnormal

structures palpable on one or more ovaries. No attempt was made to refine the diagnosis

further. Throughout the study period, a few cows (n . 20) were treated with

intramammary antibiotics on the basis of an elevated individual cow somatic cell count

recorded at last herd test. These events were grouped with mastitis events in the disease

incidence analyses.

Fig. 1. Map showing the location of the southern highlands district of New South Wales. Abbreviations: QLD

Ð Queensland, NSW Ð New South Wales, VIC Ð Victoria.

M.A. Stevenson / Preventive Veterinary Medicine 43 (2000) 1±11 3

Table 1

Definitions of disease diagnoses

Disease category Diagnosis Comments

Calving-associated disorders Dystocia-assisted delivery No malposture or foetal oversize reported

Dystocia-foetal oversize No malposture,  traction required to effect delivery

Dystocia-foetal malposture Correction of malposture required prior to delivery

Dystocia-twins Twins requiring correction in posture or assistance

to deliver

Obstetric paralysis Recumbency post-calving accompanied by a

history of dystocia and evidence of reproductive

tract trauma

Uterine prolapse Characteristic clinical signs

Metabolic disorders Milk fever Characteristic clinical signs and response to

therapy

Grass tetany Characteristic clinical signs and response to

therapy

Ketosis Positive Rothera test and exclusion of a primary

cause

Reproductive tract disorders Reproductive-tract trauma Characteristic clinical signs

Retained foetal membranes Membranes defined as retained if not passed with

24 h after calving

Uterine infections Purulent discharge on vaginal examination

Cystic ovarian disease History of abnormal oestrus cycles accompanied

by abnormal findings on ovarian palpation

Abortion Foetus/foetal membranes observed and/or oestrus

detected after positive diagnosis of pregnancy

Udder disorders Mastitis Clinical mastitis or elevated individual-cow somatic-

cell count warranting, in the herd manager’s

opinion, intramammary therapy

Teat disorders Includes teat trauma, teat peas, blind quarters, teat

lesions

Udder disorders Includes ruptured suspensory ligaments, udder

oedema

Locomotor disorders Foot disorders Any condition affecting the foot (footrot, solar

bruising, foot abscess)

Leg disorders Any condition affecting a limb proximal to the foot

(injuries, joint problems)

Other musculoskeletal

disorders

Any condition causing lameness where the primary

lesion is neither the foot or the leg (injuries,

disorders of the spine)

Miscellaneous disorders Rumen disorders Includes acute indigestion, bloat, vagal indigestion

Abomasal disorders Includes left and right abomasal displacements,

abomasal ulceration

Lower digestive-tract

disorders

Includes diarrhoea, intestinal obstruction, intestinal

volvulus

Hepatic disorders

Respiratory-tract disorders

Miscellaneous conditions Includes pinkeye, septicaemia, localised abscesses,

woody tongue, lumpy jaw

4 M.A. Stevenson / Preventive Veterinary Medicine 43 (2000) 1±11

At the start of the study period (1 January 1994), all participating managers had been

on the herd-health program for a minimum of 6 months and were familiar with the

process of maintaining a dairy and recording the appropriate disease diagnoses, if they

were encountered. Each cow that calved after 1 January 1994 and before 31 December

1995 was enrolled in the study. After 31 December 1995, each cow was monitored until it

was removed from the herd or until a calving event occurred Ð whichever occurred first.

Data recorded in DairyCHAMP was exported to the data management software Microsoft

Access for Windows Version 2.0 (Microsoft, Redmond, USA) for calculation of disease

incidence.

Incidence density (ID) was defined as the total number of episodes of disease divided

by the total number of cow-yr at risk, expressed as cases per 100 cow-yr at risk (Dohoo

et al., 1983). For reproductive disorders, the period between calving and conception (or

date of removal from the herd if the cow was not pregnant at the time of removal) was

considered to be the time at risk. For each disease, a minimum time had to elapse between

two disease episodes in order for them to be considered as two separate diagnoses (Dohoo

et al., 1983). The minimum times were: reproductive-tract infections, 2 weeks,  mastitis, 1

week,  teat disorders, 2 weeks,  lameness, 2 weeks,  digestive-tract disorders, 1 week,

respiratory-tract disorders, 2 weeks,  and abortions, 100 d.

Only one episode of each disease event was recorded per lactation for disorders

associated with calving, and for these conditions cumulative incidence (CI) was used

(Dohoo et al., 1983). For abortions, the incidence measure was cases per 100 confirmed

pregnancies.

3. Results

3.1. Farm characteristics

Approximately 17% of the state’s dairy cattle population of 91 000 head are grazed in

the southern highlands and adjoining south coast region of New South Wales. The

average herd size in this study ranged from 115 to 173 adult cows. In seven of the eight

herds, Holstein±Friesian was the only breed,  the eighth herd was made up of an equal

mixture of Holstein-Friesians and Australian Illawarra Shorthorns. All farms produced

milk that was sold under a quota system to a single dairy-herd co-operative as milk to be

used for human consumption. The daily milk quotas for these herds ranged from 1200 to

1900 l per day. Milk production was the primary source of income for all farms in this

study. In three of the herds, the sale of surplus heifer replacement stock was a (minor)

source of additional income. All farms had been established for at least 15 years and

were, in all cases, staffed by two full-time managers.

3.2. Cow husbandry

Cows grazed a pasture mixture of ryegrass and clover all year round. Pasture provided,

on average, approximately 70% of the daily energy intake. Small amounts (approximately

2±8 kg per cow per day) of grain-based concentrates fed at milking time provided the

M.A. Stevenson / Preventive Veterinary Medicine 43 (2000) 1±11 5

remainder of daily energy requirements. In times of pasture deficit, varying amounts of

hay and/or grass silage were fed as supplements. Grass silage, if it was used, was made

on-farm,  hay was purchased.

3.3. Milk harvesting

Cows were milked twice daily in herringbone milking parlours. Milk yields were

measured once monthly by the New South Wales DHI association and assisted herd

managers to make drying-off and culling decisions based on production. Managers used a

variety of measures to maintain milk quality including teat-dip application at the end of

milking and dry-cow antibiotic therapy. To identify cows with subclinical mastitis, all

managers used the California Mastitis Test in combination with the results of individual

cow somatic-cell counts measured at every second herd recording.

3.4. Breeding practices

These were non-seasonally calving dairy herds. All herds used artificial insemination

for at least the first one to three services after calving. Cows that failed to

conceive after three services were run with a bull for the duration of standing oestrus

in six herds. Artificial insemination was used as the only method of breeding in the other

two herds.

For oestrus detection, managers relied on observing their cows during milking and

grazing. Kamar heatmount detectors1 (ImmuCell, Portland, USA) were used on selected

cows. Teaser bulls were not used in any of the herds.

3.5. Replacement management

All farms reared their own heifer replacements. Calves were left with the cow for up to

24 h after birth and were then bucket-fed on whole milk. Calves were weaned at 6±8

weeks of age when they were sent to grazing (usually to an area owned by the farmer

separate from the main farm). Six of the eight farms used outdoor calf hutches to rear

unweaned calves during the study period.

3.6. Study cohort

During the study period, 1430 cows were monitored over 2111 calving events Ð

resulting in a total observation period of 745 611 cow-d (2043 cow-yr). For reproductive

disorders, the total time at risk was 291 374 cow-d (798 cow-yr). During the study period,

588 cows entered the cohort (all as primiparous cows), and 522 (25% of the total of the

total lactations started) were terminated by removal from the herd.

Descriptive statistics of 305-day milk yield (l), 305-day milk fat yield (kg), and

305-day milk protein yield are shown in Table 2. For all parity groups, 305-day milk, fat,

and protein yields were greater (by inspection) than the New South Wales DHI state

average for 1994±1995 (New South Wales Agriculture and Fisheries, 1995).

6 M.A. Stevenson / Preventive Veterinary Medicine 43 (2000) 1±11

3.7. Incidence measures

Incidence measures were calculated by diagnosis category and by individual

codes (Tables 3 and 4, respectively). Among the disease categories, calving-associated

disorders, reproductive disorders and udder disorders were the most frequent disease

events.

4. Discussion

4.1. Sources of bias

Random selection of farms for this study was not possible as a result of the continuous

co-operation that was required of the participant herd managers. Purposive selection

therefore appears to have in a set of higher-producing herds compared with the New

South Wales DHI association average for the same period. This bias may therefore limit

the generalisation of the results of this study to the target population of New South Wales

dairy herds. A further source of bias is that all of the participant herds received monthly

Table 2

Descriptions of 305-day milk, fat and protein yields by parity group in studied southern highlands dairy herds

(n . 1901 complete lactation records from eight herds in New South Wales (NSW), Australia)

Production

variable

Parity group NSW DHI

mean 1995

1 2 3±6 _7

Mean SD Mean SD Mean SD Mean SD

305-d milk (l) 5811 955 6662 1180 7365 1243 6994 1037 5736

305-d fat (kg) 222 37 252 46 279 49 263 49 215

305-d protein (kg) 188 30 218 37 239 39 227 34 182

Table 3

Between-herd variability in disease incidence in studied southern highlands dairy herds (n .

Disease category Mean SD Median Minimum Maximum

Calving-associated disordersa 18.0 11 15.3 5.0 42.3

Metabolic disordersb 5.5 1 5.2 3.9 8.7

Reproductive tract disordersc 22.3 11 19.1 8.2 38.7

Abortionsd 2.3 1 2.1 0.9 4.2

Udder disordersb 17.6 11 17.3 6.5 33.1

Locomotor disordersb 3.7 3 3.0 0.0 8.2

Miscellaneous disordersb 4.1 3 4.1 1.5 9.6

a CI: Cases per 100 calvings.

b ID: Cases per 100 cow-yr.

c Cases per 100 cow-yr open.

d Cases per 100 confirmed pregnancies.

M.A. Stevenson / Preventive Veterinary Medicine 43 (2000) 1±11 7

Table 4

Total period at risk, cumulative incidence or incidence density (and 95% CI) and time to first diagnosis (from

calving) for each of the common diseases in studied southern highlands dairy herds (n .

Disorder Total period

at risk

Incidence 95% CIa Days to first diagnosisb

Q1c Median Q3d

Calving-associated disorderse 2111 calvings

Assisted delivery 12.6 11.2±14.1 0 0 0

Foetal oversize 1.2 0.8±1.7 0 0 0

Foetal malposture 0.4 0.2±0.8 0 0 0

Retained foetal membranes 3.1 2.4±3.9 1 1 5

Obstetric paralysis 0.5 0.3±0.9 0 0 0

Uterine prolapse 0.1 0.0±0.4 0 0 0

Metabolic disordersf 745611 cow-d

Milk fever 5.2 4.0±6.7 0 0 1

Grass tetany 0.1 0.0±0.4 3 6 19

Ketosis 0.05 0.0±0.3 28 28 28

Reproductive-tract disordersg 291374 cow-d

Tract trauma 1.6 0.9±2.8 0 27 72

Uterine infections 19.9 16.9±23.3 14 34 99

Cystic ovarian disease 0.7 0.3±1.6 72 125 212

Abortionsh 1674 pregnancies 2.3 1.7±3.2 147i 174 210

Udder disordersf 745611 cow-d

Mastitis 16.4 14.7±18.3 3 37 129

Teat disorders 1.1 0.7±1.7 24 104 177

Udder disorders 0.05 0.0±0.3 25 49 74

Locomotor disordersf 745611 cow-d

Foot disorders 3.5 2.7±4.4 32 103 209

Leg disorders 0.1 0.0±0.4 19 37 165

Other musculoskeletal disorders 0.1 0.0±0.3 13 26 40

Miscellaneous disordersf 745611 cow-d

Rumen disorders 1.1 0.7±1.6 25 49 112

Abomasal disorders 0.3 0.1±0.6 26 43 138

Lower digestive tract disorders 0.5 0.3±1.0 17 59 134

Hepatic disorders 0.2 0.0±0.6 316 316 316

Respiratory tract disorders 0.1 0.0±0.4 37 70 102

Other conditions 1.3 0.8±1.9 17 81 167

Undiagnosed illness 0.5 0.2±0.9 27 50 150

a Fleiss quadratic 95% CI.

b Days from calving to first diagnosis.

c Q1: 25th percentile.

d Q3: 75th percentile.

e CI: Cases per 100 calvings.

f ID: Cases per 100 cow-yr.

g Cases per 100 cow-yr open.

h Cases per 100 confirmed pregnancies.

i Days from conception to diagnosis.

8 M.A. Stevenson / Preventive Veterinary Medicine 43 (2000) 1±11

herd-health visits and this may have resulted in higher estimates of reproductive-tract

disorders compared with those studies conducted by questionnaire only.

Bias originating from managers forgetting to record events or misclassifying events

is an obvious problem in this type of study. To provide useful data, participant herd

managers underwent a period of training to become accustomed to recording details of

all animal-health events. To assist this process, it was necessary to provide simple

and concise definitions for each of the common disease conditions,  during the training

period, it was necessary to scrutinise the herd-dairy information to clarify that each

disease event recorded matched the condition that was actually encountered. During the

training period, this process was intensive (requiring at least 30 min of attention at each

herd visit). Once managers were accustomed to the system, scrutiny of herd records was

still necessary although the time taken to complete the process was less, at the most

10 min at each herd visit. Regular contact with the participant farmers, regular feedback

of their herd performance and anonymous comparison of reproductive performance,

production, and disease incidence levels with their peers maintained interest throughout

the period of study and enabled (what I believe to be) a high level of data quality to be

obtained.

4.2. Disease incidence

During data entry into the DairyCHAMP program, it was necessary to review the

disease-event history of each cow to ensure that the necessary time had elapsed between

similar disease events (that is, to account for a recovery period). An improvement to this

software package would be a warning to the user (at the time of data entry) that a second

disease diagnosis is being entered within a pre-defined recovery period. No detailed

documentation of DairyCHAMP’s method of calculating disease incidence was available

and because of this it was necessary to export the raw event data into a data-management

program. To facilitate further studies of this nature, it would be helpful if consensus was

reached regarding the calculation of incidence for the various dairy cow diseases and that

these methodologies be transparently applied to dairy-cow herd-health software

packages. If this situation eventuated (and provided that similar disease definitions were

used), dairy-cow disease-incidence studies drawn from wider geographic areas and using

greater numbers of herds would be possible.

The incidence of each of the major disease categories varied across herds with wide

variations recorded for calving-associated disorders, udder disorders and reproductive

tract disorders (Table 3). With respect to calving-associated disorders, a large part of the

between-herd variation was accounted for by those calvings where delivery was assisted

in the absence of any foetal oversize or malpostureÐa finding that reflects differences in

the intensity of husbandry provided by each of the participant herd managers. Variation in

the incidence of udder disorders and reproductive-tract disorders between herds reflects

differences in managerial styles and skill at controlling each of these conditions (Faye,

1991).

A comparison of the results reported here with other published data is shown in Table

5. Care must be taken in making detailed comparisons between disease-incidence studies

in dairy cattle because differences exist in the health conditions that have been recorded,

M.A. Stevenson / Preventive Veterinary Medicine 43 (2000) 1±11 9

the criteria used to define health disorders and the way disease incidence has been

reported. Although the magnitude of incidence varied between studies, the relative

importance of disease categories was similar with reproductive-tract and udder disorders

being the most important. While the overall incidence of reproductive disorders is low in

this study compared to others (Dohoo et al., 1983,  GroÈhn et al., 1989,  Frei et al., 1997), it

should be appreciated that only cases of identifiable reproductive-tract pathology were

recorded Ð ignoring events such as cows presented for anoestrus post-calving and for

failure to conceive.

5. Conclusion

Detailed monitoring of disease events and animal treatments was possible in this group

of herds but required commitment on the part of the herd manager, a 3±6 month training

period to become familiar with the recording system and close and regular contact with

the herd veterinarian. Given these strict provisions, it is my belief that the system of

disease classification used in this study is suitable for dairy-herd managers wishing to

provide complete documentation of their herd’s disease status.

The software used to record animal-event, disease and production data was well-suited

for this type of study (although a more comprehensive description of the method used to

calculate disease incidence within the package would have been helpful). To facilitate

larger-scale studies of dairy-cow disease incidence, definitions of animal disease need to

be agreed upon across data-collection centres and similarly, disease-incidence

calculations need to be standardised.

In agreement with dairy-cow disease-incidence studies conducted in other parts of the

world, disorders of the reproductive tract and udder were the most frequent clinical

conditions encountered. These findings confirm the conclusions of other studies that

dairy herd-health programs should be emphasise the control of these two groups of

disorders.

Table 5

Comparison of disease incidence in dairy cows (ID per 100 animal-years)

Disease category Switzerlanda Michiganb Australiac This study

Calving-associated disorders 8.9 13.8 NRd 18.0

Metabolic disorders 11.6 10.2 3.9 5.5

Reproductive tract disorders 152.2e 49.9 14.0 22.3

Udder disorders 40.5 33.1 15.3 17.6

Locomotor disorders 16.4 6.6 NR 3.7

Miscellaneous disorders NR NR 10.2 4.1

a Frei et al. (1997).

b Kaneene and Hurd (1990).

c Williamson et al. (1978).

d NR: not reported.

e Included non-clinically apparent reproductive tract disorders such as non-visible oestrus.

10 M.A. Stevenson / Preventive Veterinary Medicine 43 (2000) 1±11

Acknowledgements

The author gratefully acknowledges the assistance provided by the participant farmers

during the course of this study and the staff of W.R. Beresford and Associates, 472 Argyle

St. Moss Vale, New South Wales, Australia. The comments and advice of Dr Ron Jackson

are greatly appreciated.