| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,1
,
* Department of The Norwegian Cattle Health Services, TINE Norwegian Dairies BA, 1431 Ås, Norway
Department of Production Animals Clinical Sciences, Norwegian School of Veterinary Science, 0033 Oslo, Norway
GENO Breeding and AI Association, 2326 Hamar, Norway
Norwegian Quality System for Agriculture, 0135 Oslo, Norway
|| Department of The Norwegian Dairy Herd Recording System, TINE Norwegian Dairies BA, 1431 Ås, Norway
# Norwegian Veterinary Association, 7710 Sparbu, Norway
1 Corresponding author: olav.osteras{at}veths.no
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: recording system • bovine health • disease recording
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In some countries, systematic epidemiological surveys of disease incidences in dairy production have been organized, such as the National Animal Health Monitoring System in the United States (Kaneene and Hurd, 1990) and others in Canada (Sargeant et al., 1998; van Dorp et al., 1999). However, only a few countries have reported disease recordings from the majority of the dairy cattle population within the framework of an animal recording system. These countries are Norway (Solbu, 1983), Sweden (Emanuelsson and Oltanacu, 1998), Finland (Saloniemi and Roine, 1981), and Denmark (Bartlett et al., 2001). Norway began its disease recording system in 1975–1976. Solbu (1983, 1984a,b) has previously published some of these data.
The intention of this paper is to present the organization of the Norwegian Cattle Health Recording System (NCHRS) in the past and at present, the experiences with 2 different coding systems, and an evaluation of the results obtained during a 30-yr span from 1975–1976 through 2006, and further to discuss these results with special reference to present and future use of the data.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
|
Reporting and Structure on Database
Periodically (once a month), the records from the Cow Health Cards are reported by the Dairy Herd Improvement Services into the mainframe of the NDHRS. The complete structure of these data tables and more detailed information is available at www.kkvet.no. The data are processed annually to create the official annual reports of the NDHRS and the NCHRS (Norwegian Dairy Herd Recording System, 1976–2006).
The Norwegian Dairy Association joined the formal agreement of the Norwegian Cattle Health Card System in 1979, and in 1989 the official Veterinary Authorities also joined. In this way, data from the Norwegian Cattle Health Card System from 1989 became part of the official disease reporting system for dairy cattle in Norway. From 1996 the responsibility of coordinating the disease codes for production diseases was turned over from the official authorities to the Norwegian Cattle Health Services (NCHS). For more information, consult http://storfehelse.tine.no. From the start of the recordings, the diseases expected to be most prevalent were given a unique number from 1 to 43. The code numbers from 50 to 56 were introduced in 1979 (see Table 2). In 1989 the coding system was totally rearranged to an organ-related coding system (Table 2), in which each 100 represents diseases related to separate organ systems. A coding system for the preventive therapy of each disease was achieved by adding 500 to the disease code. Thus, code 766 is a preventive treatment of disease code 266, parasitic disease. From August 1999 a few new disease codes on reproduction and udder health (Table 2) were introduced and, at the same time, the name of acute clinical mastitis was changed to severe and moderate clinical mastitis, and chronic clinical mastitis was changed to mild clinical mastitis in accordance with the recommendations of the International Dairy Federation (1997). This paper deals only with the results from dairy cattle with records from first calving until removal from the herd.
|
The incidence rates of diseases are calculated according to Martin et al. (1987) and Kleinbaum et al. (1982). The incidence rate (incidence density) is defined as the number of cows with one or more events of disease treatment divided by the number of cow-years at risk within one calendar year (COW_RATE). According to the International Dairy Federation recommendations on presentation of mastitis-related data (International Dairy Federation, 1997), the number of such events is defined as the number of cows with the event during a specified year. The number of cow-years at risk is defined as the sum of the total number of days for each cow within the herd during the specific year, adding up all days from the first day at first calving until the culling day, all within the specified year. A new case of a diseased cow will be calculated despite the occurrence of the same event in any earlier year. The number of cow-years at risk is corrected in accordance with International Dairy Federation (1997) recommendations. This is based on the assumption that, on average, during any one year, the event happens halfway through the lifetime of the cow that year. The mean lifetime of the cows was calculated by dividing the total number of cow-years by the total number of cows in the population during the year in question. Thus, each diseased cow is not at risk during half the mean number of days that cow represents when calculating the total cow-years in the current year. The formula for this calculation is:
 |
where N_COW_TREAT is the number of cows having at least one case of the actual disease, COW_YEARS is the total number of cow-years within the whole population during the year in question, and N_COW is the total number of cows in the population during the year in question.
For the 10 diseases most frequently encountered throughout the complete period, the treatment incidence rate (TREAT_RATE) was calculated. The TREAT_RATE was defined as the number of recorded treatments divided by the total number of cow-years within the population; thus:
 |
where N_TREAT is the number of recorded treatments for the actual disease code during one year, and COW_YEARS is the total number of cow-years within the whole population during the specified year. The variation in COW_RATE between years was estimated as the standard deviation of the COW_RATE each year, also divided by the overall mean COW_RATE multiplied by 100, or percent standard deviation.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The mean milk yield produced per cow-year increased in the population from 5,428 kg in 1975 to 6,403 kg in 1993. From 1993 until 2000, the milk production per cow-year was considered stable because of the milk quota system. Later, it increased again to a level of 6,605 kg per cow-year until 2005. The composition of the feed consisted mainly of grass silage (28.4% of the energy content in 1975 to 41.0% in 2005), concentrate (38.6% in 1975 to 36.8% in 2000, up to 39.9% in 2005), and pasture (22.9% in 1975 to 13.9% in 1992, up to 16.7 % in 2002). Other minor feed components were hay (5.5% in 1975 to 0.5% in 2005) and roots (3.4% in 1975 to 0.05% in 2005). Other feedstuffs constituted 1.2% of total feed in 1995 and up to 7.3% in 1993.
Altogether, 4,444,485 cows were recorded with one or more diseases during the 30 yr of health recording of Norwegian dairy cattle. A total of 8,435,704 disease cases were observed in a base population of 8,633,532 cow-years and 8,632,516 calvings. A total of 11,563,692 individual dairy cows were represented for one or more days during the 30 yr. This represents the largest database of continuous recording of disease for dairy cattle that we are aware of. The incidence rate for all diseases, cows treated for any disease, estimated incidence risk (1 – e–COW_RATE) for any disease, and number of recorded treatments per treated cow each year from 1975 through 2005 is presented in Figure 1.
|
, as well as the incidence rate (COW_RATE) every fifth year. Overall, the most frequent disease recordings were for acute (severe and moderate) clinical mastitis (21.28 ± 3.99 per 100 cow-years); ketosis (13.83 ± 6.42); chronic (mild) clinical mastitis (9.40 ± 4.70), milk fever (6.54 ± 1.47), and teat tramps (4.15 ± 1.22); retained placenta (3.99 ± 0.65); absence of heat (3.43 ± 1.43); cystic ovaries (1.74 ± 0.38); metritis (1.59 ± 0.39); and dystocia (1.24 ± 0.42). The COW_RATE for these diseases is illustrated in Figures 2 to 5.
|
|
|
|
The most stable disease COW_RATE from year to year, expressed as the percentage standard deviation compared with the overall mean, was for the following disease records: retained placenta (16%), heat synchronization (18%), paresis caused by pelvic and or muscular injury (19%), severe and moderate clinical mastitis (19%), abscesses and phlegmons (20%), arthritis (21%), parasitic disease (22%), cystic ovaries (22%), milk fever (23%), and abortion (24%).
Number of Treatments per Diseased Cow
Table 3 presents the TREAT_RATE and Table 4 presents the number of treatments per treated cow of the most prevalent diseases from 1994 to 2005, such as clinical mastitis, ketosis, milk fever, teat tramps, absence of heat, and retained placenta. The number of treatments per treated cow showed a decreasing trend from 1994 through 2005 for all diseases. Ten diseases had less than 1.06 treatments per treated cow, on average, during the period 1994 through 2005. These diseases were dystocia, retained placenta, prolonged gestation, torsion of the uterus, decornuation caused by injury, abortion, malformation, fractures, uterine prolapse, and extirpation of supernumerary mammary teats. Ten diseases had more than 1.25 treatments per treated cow, on average, during the period 1994 to 2005. These diseases were arthritis (1.38), milk fever (1.38), ketosis (1.36), abscesses and phlegmons (1.36), diseases of the respiratory tract (1.36), acute (severe) clinical mastitis (1.34), deficit of vitamins or minerals (1.30), hoof diseases (1.28), malignant catarrhal fever (1.26), and colic, gastrointestinal dislocation, and dislocation of the abomasum (1.26).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Study Population
The study population comprised all herds within the NCHRS and NDHRS. The disease incidence for the whole country would be biased if the incidence differed between those herds that are not members and those that are. The proportion of cows that are members is greater than the proportion of herds. This indicates that herds outside the system have relatively fewer cows. The annual reports (Norwegian Dairy Herd Recording System, 1976–2006) indicate that the incidence of disease treatment is higher in smaller herds than in larger herds under Norwegian conditions. This means that the underestimation of incidence rates for the total Norwegian cattle population is more pronounced for the 1970s than for the 1990s. It is obvious that the relative value of one single cow, compared with the value of the whole herd, is higher in smaller herds than in larger herds. The farmer will therefore have a higher investment interest in treatment to keep the cow within smaller herds. This could have the effect that the treatment rate is higher, assuming that the true disease rate is the same.
Incidence Rates
The different calculation techniques can cause a difference in disease rate by more than 30% within the same population (International Dairy Federation, 1997). The main goal of this study was to present incidence trends in the Norwegian population of dairy cattle throughout 30 yr, year by year. The data available were the number of cows diagnosed with any disease according to the coding system presented in Table 2 for each year, and the base population with the number of cow-years observed from first calving to slaughter or removal within the period of January 1 through December 31 each year. This is why some treatments that rationally occur only once during a lactation, such as dystocia, still show a repeated treatment per diseased cow of 1.02. A number greater than 1 indicates that 2 calvings could occur within 1 yr of observation. The number of cow-years at risk was the closest estimate to the true incidence rate available and will be correct, assuming that, on average, the disease event would occur halfway between calving and the end of lactation.
The estimated treatment rate will reflect the frequency of treatments per cow-time. For a population this will be the true rate of therapy for a specific disease in the population in any cow, whether treated earlier or not. For diseases that are not very common, the treatment rate and incidence rate would be almost equal.
Most Common Diseases Compared with Other Surveys.
Clinical mastitis is also the most common disease in other studies in the literature that present disease rates. A study from France revealed that the most common disease in dairy cattle was clinical mastitis (with 44.1 cases per 100 calvings), locomotor disorders (with 10.9 cases), digestive disorders (with 5.1 cases), retained placenta (with 8.8 cases), dystocia (with 6.6 cases), milk fever (with 5.6 cases), and chronic metritis (with 5.1 cases; Fourichon et al., 2001). The largest difference compared with our study was the difference in locomotor disorders. This could be because Norway has relatively few claw problems in tie-stall systems and very few cases of infectious digital dermatitis (Sogstad et al. 2005). A study from Denmark revealed an incidence rate of clinical mastitis of 36 to 48 cases per 100 cow-years at risk (Bartlett et al., 2001). A British study covering 340 herds had 36.6% mastitis, 23.7% lameness, 5.3% hypocalcemia, 8.7% assisted calving, 1.3% digestive diseases, 0.7% hypomagnesemia, and 0.4% ketosis (Whitaker et al., 2000). The 5 most frequent diseases in Finland were acute mastitis (17.0%), anestrus (8.1%), ovarian cysts (7.3%), milk fever (5.4%), and ketosis (4.9%), according to Rajala and Gröhn (1998). The mastitis incidence rates from different countries vary from 0.14 cases per 365 cow-days through 0.48 cases per 365 cow-days (Barkema et al., 1998; Sargeant et al., 1998; Beaudeau et al., 2002; Peeler et al., 2002). During the period under consideration in the present paper, the same variation within the Norwegian population was estimated to be between 0.50 in 1994 and 0.25 cases of clinical mastitis per 365 cow-days at risk. This illustrates similar variation within Norway during 10 yr as was reported from different countries during recent years. This is important, because it illustrates that an accurate quantification of disease incidence cannot be made simply by monitoring disease over a single year.
Number of Treatments per Treated Cow.
Some of the diseases that are very frequent, such as ketosis and mastitis, may occur several times during the lactation. Other diseases are more infrequent, but when they occur they are very difficult to treat. Examples of these hard-to-treat diseases are arthritis, phlegmons, respiratory diseases, hoof diseases, and malignant catarrhal fever. The reported incidence of malignant catarrhal fever in this study was 0.1 per 1,000 animals, whereas from Switzerland an incidence of 0.6 per 1,000 was reported (Muller-Dublies et al., 2001). These numbers are probably underestimated, because the Norwegian data in this paper consisted only of milking cows and not young stock.
Diseases with a low number of treatments per diseased cow were those that respond very well to therapy or those that, when veterinarians are called, mean there is a problem that has to be solved at once. For some of these diseases, it is typical that if they cannot be cured, the animal will be slaughtered.
Effect of Introducing New Codes.
When new disease codes were introduced, as in 1979, 1989, and 1999, it would typically take about 5 yr before the new coding system was functioning well in the field. This was especially true for the differentiation of the clinical mastitis codes in 1979. In 1989, when the whole coding system was extensively changed, all the health cards were centrally preprinted with the cow identity and all the previously reported codes for that cow. This ensured the quality of results when this large change was performed.
Trends in the Most Common Diseases.
The increase in clinical mastitis from the 1980s is probably because of increased awareness on the part of farmers of the need to keep the bulk milk SCC at a low level to satisfy the requirements of the quality payment system. Typically, there has been an increase in treatment of chronic (mild) cases. The decrease from 1994 onward has 3 main causes. One is an action by the producers’ organization to reduce the unnecessary use of antibiotics. The second is the effect of the breeding program, as documented by Heringstad et al. (2003), who demonstrated a genetic improvement in the Norwegian cattle population from 1990 onward. The third cause is the establishment of NCHS in 1995 and a good reporting system with available data that can be used as a production management tool.
The other major disease change is the increase in ketosis, with an incidence rate at 0.14 per cow-year in 1980, an increase to 0.24 per cow-year in 1985, and then a decrease from the mid-1980s to a relative low incidence of 0.04 in 2005. The increase is probably a result of the management and feeding strategy when the quota system was introduced in January 1982. Under the quota situation, it is a goal to produce a certain amount of milk as cheaply as possible. In this situation, some farmers may have reduced the feeding of concentrates. The reduction of ketosis in recent years is probably caused by a greater awareness of and concern about the ketosis problem and subsequent adjustment to feeding and the feeding strategy from the 1990s onward. Cow have more free access to high-quality roughage and a better distribution of concentrates during the day. Finally, during 2005, the incidence rate of ketosis was lower than for milk fever, which is also typical for other countries (Fourichon et al., 2001).
The decrease in teat tramps from 1989 could partly be due to the focus on milking management and pulsator function after it was revealed that the milking process and the milking machine were the most important risk factors for teat tramps (Østerås et al., 1990; Østerås et al., 1995).
The increase in treatment of anestrus from 1980 could be caused by access to new medicines, such as prostaglandins to treat silent heat. The decrease from the middle of 1980s could be linked to the decrease in ketosis because of the positive relationship between ketosis and reproductive diseases such as anestrus and cystic ovaries (Refsdal, 1982; Fourichon et al., 2000). The decrease in therapy for retained placenta could be caused by the change in treatment strategy from active removal to the more common nonremoval treatment strategy practiced from the 1990s onward. Research has shown little difference between those 2 strategies (Drillich et al., 2003).
Variation in Disease Treatments.
The incidence of respiratory disease showed marked spikes in 1976, 1990, and 1995 (Figure 6). This could be because of well-known epidemic outbreaks of bovine respiratory syncytial virus in parts of Norway during those years (Norström et al., 2001). After each peak of disease in the respiratory tract, the general level was established at a higher level than before. This could be because a new strain of bovine respiratory syncytial virus was introduced, as probably was the case in 1995 when the introduction was caused by an import of live animals from Denmark (Norström et al., 2001). The increased treatment of louse infestation with the introduction of eradication programs (Nafstad and Grønstøl, 2001) fits well with the experience that lice and external parasites have been more common and more commonly treated in recent years.
|
) could be linked to the Chernobyl nuclear power plant catastrophe in the last week of April 1986. The peak occurred 3 to 5 yr after the fallout of radionuclides Cs134 and Cs137. There is also a distinct but smaller peak in the year 1986. These 2 peaks are significant, and there could be 2 reasons for this finding. Either there was more attention to malformations after Chernobyl, or there was a true increase in malformations because of the biological effects of radionuclides deposited over certain areas of Norway. More recently, the incidence of this diagnosis has reduced continuously to a lower level than ever before. In the Norwegian human population, an association was identified between total radiation dose and hydrocephaly. No associations were found, however, for conditions previously reported to be associated with radiation, namely, small head circumference, congenital cataracts, anencephaly, spina bifida, and low birth weight (Lie et al., 1992). The decrease in diagnosis and therapy of narrow teats from 1980 through 1988 could be because of genetic selection for milkability. Milkability has a high heritability of 0.37 ± 0.12 (Miller et al., 1976), and the trait was introduced into the breeding program both at the sire level (national level) and at the dam level (herd-level selection). The trait reached such a low level that the diagnosis was removed from the recording system.
Validity
When diseases are recorded under practical farming conditions, there will always be some reasons for mis-classification. Such misclassification could be because of simple errors in writing down individual numbers, diagnosis codes, and treatment dates. The errors could occur at the veterinary level, the farmer level, or the reporter level. Some of these errors could be identified as diagnoses in males that usually occur only in females, or diagnoses before calving for those that should only occur after calving. Some errors will always be hidden within the large number of other relevant diagnoses. Such errors will usually occur at a higher proportional rate for those diagnoses that are very infrequent compared with the more common diagnoses. Accordingly, our experience has been that figures on events that are diagnosed less than 50 times per year should be judged and evaluated with great care. The figures from Table 2 for diseases such as tetanus, actinomycosis, and kidney disease were therefore removed from the coding system in 1989. Currently, the Norwegian disease code for production animals includes 67 different diagnoses. Our experience is that there is no need for additional differentiation because diseases with low frequency are not robust. The recording system will be used mainly for daily management at the farm level, survey of the disease situation, and animal breeding. If there is a need for research on more infrequently occurring diseases, then the research should be designed in a way that is different from general surveys. The organ-related system makes it easy to add new diagnosis codes, as was done in 1999, if there are available codes under that organ system. The system is thus robust for certain changes.
At present, the NCHS is working on different evaluation processes for such data. There has been a comparison of retail sales of intramammary drugs used for mastitis therapy and the incidence rate of clinical mastitis throughout the 30 yr. These comparisons revealed very parallel curves for incidence and retail sales. This indicates that even if some data are missing, the same proportion is missing every year. This ensures that the reduction in mastitis treatment recorded from 1994 through 2003 is a true reduction, because it is also reflected in the retail sales of intramammary drugs for mastitis therapy.
There is an urgent need for good validation processes as the need for health records increases for daily farm management, breeding purposes, and traceability. More research should be done to evaluate the performance in practice of such validations.
The organ-related code system from 1989 made it easier to identify a certain disease code. A precise scientific diagnostic code system could very easily be too detailed and be a source of more misclassification. The NCHRS is completely integrated into daily practical farming needs and is incorporated into the national breeding program, which is also driven by farmers.
The experience with NCHRS for dairy cattle has shown some promising results, because it has extended awareness of the high incidence rates for both mastitis and ketosis throughout the 1980s. Since 1994 it has been the most important tool to decrease the overall treatment rate by more than 50% by means of management tools, information processes, periodical reports, and breeding programs. The continuous recording for several diseases also demonstrates a huge variation over the years, illustrating that a national survey for 1 yr would not reliably indicate the level 5 yr later. One example is that both mastitis and ketosis declined considerably in the course of a few years in Norway, mastitis by 52% and ketosis by 75%. Such reductions would be impossible without knowing the incidence rate in each year and, on this basis, setting goals and evaluating progress. Access to the data is also essential for breeding purposes and scientific research to be able to make further improvements.
Received for publication January 14, 2007. Accepted for publication May 28, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. T. Garmo, A. D. Martin, E. Thuen, O. Havrevoll, H. Steinshamn, E. Prestlokken, A. Randby, M. Eknaes, A. Waldmann, and O. Reksen Characterization of progesterone profiles in fall-calving Norwegian Red cows J Dairy Sci, October 1, 2009; 92(10): 4919 - 4928. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Gulliksen, E. Jor, K. I. Lie, I. S. Hamnes, T. Loken, J. Akerstedt, and O. Osteras Enteropathogens and risk factors for diarrhea in Norwegian dairy calves J Dairy Sci, October 1, 2009; 92(10): 5057 - 5066. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Gulliksen, E. Jor, K. I. Lie, T. Loken, J. Akerstedt, and O. Osteras Respiratory infections in Norwegian dairy calves J Dairy Sci, October 1, 2009; 92(10): 5139 - 5146. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Hovinen, M. D. Rasmussen, and S. Pyorala Udder health of cows changing from tie stalls or free stalls with conventional milking to free stalls with either conventional or automatic milking J Dairy Sci, August 1, 2009; 92(8): 3696 - 3703. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Gulliksen, K. I. Lie, T. Loken, and O. Osteras Calf mortality in Norwegian dairy herds J Dairy Sci, June 1, 2009; 92(6): 2782 - 2795. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Gulliksen, K. I. Lie, and O. Osteras Calf health monitoring in Norwegian dairy herds J Dairy Sci, April 1, 2009; 92(4): 1660 - 1669. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Walsh, F. Buckley, K. Pierce, N. Byrne, J. Patton, and P. Dillon Effects of Breed and Feeding System on Milk Production, Body Weight, Body Condition Score, Reproductive Performance, and Postpartum Ovarian Function J Dairy Sci, November 1, 2008; 91(11): 4401 - 4413. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. T. Garmo, A. O. Refsdal, K. Karlberg, E. Ropstad, A. Waldmann, J. F. Beckers, and O. Reksen Pregnancy Incidence in Norwegian Red Cows Using Nonreturn to Estrus, Rectal Palpation, Pregnancy-Associated Glycoproteins, and Progesterone J Dairy Sci, August 1, 2008; 91(8): 3025 - 3033. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Gulliksen, K. I. Lie, L. Solverod, and O. Osteras Risk Factors Associated with Colostrum Quality in Norwegian Dairy Cows J Dairy Sci, February 1, 2008; 91(2): 704 - 712. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
