Field Investigations of Endemic Diseases | Animals Veterinary Medicine

In this article we will discuss about the method of field investigation for endemic diseases seen in animals.

Epidemiology and Health Management:

Health management, as the name implies, is the action of managing the health (including prevention and treatment of disease) of animal popu­lations. In farm animals, the process represents an extension to what are currently called herd health programs. Some have coined the term planned animal health and production services (PAHAPS) for these activities.

Health management programs require knowl­edge from a number of areas, including traditional medicine (etiology, pathogenesis, diagnosis, and treatment of disease), animal behavior, nutri­tion, animal management, and housing, as well as epidemiology and eco­nomics. (One might also add selected skills from sociology and psychology, since an understanding of the owner/manager may prove vital to the intro­duction and continued success of health management programs.)

In gen­eral, health management programs are targeted at animal populations; however, the actual delivery will likely involve different levels of organiza­tion from the individual (animal/animal owner) to larger groups (herds, kennels) as the units of concern.

The specific roles of epidemiology and economics in health manage­ment programs are still evolving, but they tend to function as integrative disciplines in that they provide the concepts and tools to under­stand and investigate relationships among the factors contributing to the productivity of the animal population(s) of concern.

Although the princi­ples of health management apply to veterinary public health, private food animal and companion animal medicine, and regulatory (public) veterinary medicine, nowhere is the need for epidemiologic input greater than in the field of health management of farm animals, particularly those animals reared under intensive management conditions.

Schwabe indicate quite correctly that the current intensification of animal agriculture in North America has been made possible largely because of the efforts of publicly employed veterinarians who were able to control diseases such as Texas fever, Trichinella spiralis, contagious bovine pleuropneumo­nia, and more recently hog cholera.

Today, the national veterinary service in most countries with intensive agricultural industries has the responsibility for the ongoing exclusion of many potentially devastating diseases such as foot-and-mouth disease and African swine fever, as well as pursuing the control and/or eradication of endemic diseases such as brucellosis and tu­berculosis. All these activities are essential to provide an umbrella of pro­tection over the intensive domestic animal industries.

Epidemiologic methods were essential to these early activities in do­mestic control and still play a central role in the programs of organized veterinary medicine. The major intent of this and the subsequent section is to demonstrate and reinforce the potential value of an epidemiologic ap­proach to health management at the farm/veterinary practice level by private practitioners.

This section could begin with an exhaustive list of diseases for which the natural history remains unclear. This list would certainly include dis­eases such as bovine virus diarrhea, infectious bovine rhinotracheitis, avian mycoplasma infections, bluetongue, and Aujeszky’s disease. However, such a listing might in itself suggest that an agent by agent or disease by disease approach to disease control is the best way of proceeding.

Certainly past successes have shown that such an approach works; yet, the major prob­lems confronting domestic animal industries today are multi-etiologic in nature. Hence, a manifestational rather than an etiologic classification of problems seems more appropriate. (Multi-etiologic implies that many agents and/or many factors in addition to specific agents are involved in causing that disease.)

These multi-etiological manifestational syndromes include res­piratory disease in the swine, beef, and poultry industries, neonatal mortal­ity and reproductive inefficiencies in all species, and metabolic diseases and mastitis in dairy cows.

By their very nature, these diseases are difficult to study under controlled laboratory conditions; hence, the real world (i.e., the feedlot, swine barn, or poultry house) will become an important “labo­ratory” for their investigation. It is here that the applied techniques of epidemiology, including analytic studies, field experiments, and simulation modeling, will prove extremely useful.

It would be false to suggest that well-designed field studies have ap­peared only recently, or that without formal epidemiologic training, good field studies and field investigations are not possible. Certainly, qualitative epidemiologic skills have been used for many years, often in conjunction with microbiologic and clinical skills.

What is true, however, is that quanti­tative epidemiologic techniques have only recently been applied to investigations of problems in farm animal industries. For example, the first for­mal case-control study in farm animals was an investigation of the etiology of left displacement of the abomasum in dairy cows reported in 1968.

In domestic animals, in addition to untangling the various diseases involved in these multi-etiologic syndromes, the major questions to be re­solved are the impact of these syndromes on productivity, and identifying the factors causing the syndromes. As well as the obvious value to the animal owner, answers to these questions should provide a rational basis for establishing research priorities.

To ensure that production is emphasized as the end point, it might be instructive to identify specific deficit areas of production and then identify the causes of these deficits. It is quite likely that management errors and subclinical problems as well as clinical disease per se will be identified in this manner.

Identifying the causes of these production deficits will frequently lead to studying the interrelationships among diseases, identifying important host characteristics, and elucidating the more important environmental determinants of the problem.

Just as infectious agents affect each other directly and indirectly, and the effects of multiple infections on the host may be additive or interactive, diseases also tend to be associated with each other and their combined effects on each other and on production may be additive or interactive.

New and more exacting epidemiologic techniques applicable to health management will be developed as studies at the individual animal level progress to studies at the herd level. For example, in 1975, epidemiologic studies at the Ontario Veterinary College (OVC) were initiated into the interrelationships among diseases and their effects on productivity in 18 dairy herds.

The data base was assembled in a manual fashion by copying the information from individual cow cards, OVC hospital records, and Record of Performance production testing program records. Much data were discarded because of apparent errors, and the definitions of many of the disease syndromes had to be quite general.

A number of clinicians had input data into the medical records or on the cow cards. Consequently the diagnoses, although probably of high quality, were based on non-standardized terminology. Despite these difficulties much useful information was obtained from these initial studies.

Subsequently, a prospective study was initiated that included more herds (n = 32) in a wider geographic area serviced by three different veteri­nary practices. In this study, dairy farmers were asked to maintain records specifically for the senior investigator. In most instances this only required increased vigilance on the part of the farmer because most already had a recordkeeping system; the new feature was that someone was going to formally analyze the data.

Through regular farm visits by the senior investi­gator and with the help of the enthusiastic dairy farmers and their veteri­narians, a large high quality data base was established. Again, however, many diagnostic categories had to remain general to take account of the variation in terminology and procedures among veterinary practices.

Much useful research data were obtained in this study, and new epidemiologic techniques for case-control studies were developed to assist in its analysis. In addition, practical advice about the advisability of selected management practices (e.g., the effect of delaying the first breeding to approximately 90 days postpartum) based on formally analyzed field data was generated. Also in this study, initial attempts at explaining herd-to-herd variation in production and disease rates were completed.

The most recent epidemiologic studies at the individual cow level were based on data resulting from a field trial designed to study the efficacy of two biologics on reproductive performance. The study took place in one large (300 cows) dairy herd, and the majority of observations were made and recorded by one veterinarian.

Together with much attention to detail, this provided a high quality data base that in addition to meeting the field-trial objectives has been used to study interrelationships among diseases and their effects on productivity in dairy cows. Not only are the diagnostic criteria well defined, some of the diagnoses are supplemented by the results of laboratory tests (e.g., plasma progesterone levels).

As the use of computers in the livestock industries increases, large, accurate data bases will become available on which to base research activi­ties and from which invaluable data for extension activities can be drawn.

As dairy farmers gain positive results by keeping and analyzing (in conjunc­tion with the veterinarian and extension personnel) data on their animals, there will be a natural tendency to increase the quality and the quantity of the data recorded. Thus, future large-scale research projects may be based on data derived from recording systems primarily instigated to assist the farmer and the veterinarian to make better management decisions.

With some concerted efforts toward standardization of diagnostic terminology, such a data base, when supplemented by well-planned meta­bolic and microbiologic profiles, should allow a comprehensive picture of relationships among management factors, agents, disease, and production at the individual cow level.

It should also prove useful for studies of the association between genotype and disease occurrence. The health management area requiring increased study over the next few decades is at the herd level (i.e., the identification of factors that influence herd-to-herd variation in productivity and disease occurrence).

Just as it is difficult to understand how individuals function by examining cells and organs, it is difficult to understand how herds or other aggregates of individuals function by studying only individuals.

Until recently, however, the technology to study sufficiently large numbers of herds has not been available; the widespread use of computers and the increased avail­ability of appropriate software has largely circumvented this limitation. For example, a further major epidemiologic project involving the dairy industry and workers at the OVC focused on a random sample of southwestern Ontario dairy herds.

The 104 farms took part in a 3-year study designed to investigate associations among disease, drug usage, and productivity. Two-thirds of the farms provided farm-level data only (e.g., the number of cows with retained fetal membranes and/or metritis each month), whereas one- third provided both individual cow level and herd data (i.e., which cows had metritis).

One recent example of a health study where an aggregate of individ­uals was the unit of concern is the Bruce County Beet Health Project conducted in Ontario, Canada. This project com­menced in 1978 and continued for 3 years.

In each of the years, between 60 and 70 feedlot operators collaborated in the project by providing daily treatment and death loss records, weekly ration content descriptions, and a record of all processing (vaccinations, deworming, castration, etc.) for each identifiable group of calves.

Each year there were approximately 110 groups of cattle, containing an average of 140 beef calves each. The demo­graphic characteristics of each group of calves, their source, and method of transportation to the feedlot, as well as their housing and management were recorded by the investigators shortly after arrival. Approximately 80% of all animals that died were examined by pathologists, microbiolo­gists, and parasitologists at the OVC.

The majority of the calves in this study were highly stressed; they were raised on open pastures in western Canada, weaned, trucked to sales-yards, and shortly thereafter transported by truck or train for a period of 3-7 days (2000-3000 km) to Ontario. Some went directly to feedlots, others were sorted into homogenous groups and resold at sales-yards in Ontario.

Most of the calves had never eaten from a feed bunk or drunk water from a bowl or trough prior to this. Not surprisingly, because of these stresses and the often inclement weather during this time of the year, the calves were susceptible to many disease conditions; particularly respiratory disease, the main clinical condition being a respiratory syndrome associated with fibrin­ous pneumonia. However, because it is difficult to clinically distinguish among the respiratory diseases, the general syndrome is usually referred to as the shipping fever complex.

The findings of the pathologists reinforced the overall importance of respiratory disease with the proportional mortality rate for respiratory dis­ease varying from 54% to 64%. Yet, the proportional mortality rate for fibrinous pneumonia decreased dramatically in the last year of the study from 43% to 29% in the face of a stable overall mortality rate.

It was postulated that this decline was due to producers avoiding certain manage­ment practices that had been associated with fibrinous pneumonia in the previous years. Since it was not possible to derive accurate cause-specific morbidity data, in one series of analyses the groups of calves were catego­rized in a case-control manner into those having one or more deaths from a specific cause versus no deaths from that cause.

Differences between these groups in terms of demographic characteristics, housing, feeding, and processing factors were studied. In general, the important factors were those associated with crude mortality rates; this may have been due to the overwhelming importance of a few diseases, such as fibrinous pneumonia, bronchial pneumonia, interstitial pneumonia, infec­tious bovine rhinotracheitis, and infectious thromboembolic meningoen­cephalitis.

The major method of analysis used to sort through the large number of potential risk factors was multiple regressions. This technique allows the investigator to examine the effects of one factor while other factors in the regression equation are held constant mathematically. In this regard, least squares multiple regressions is analogous to the Mantel-Haenszel technique and is appropriate when the outcome (dependent variable) is a quantitative variable.

Logistic regression, a powerful extension of the Mantel-Haenszel technique, also was used in one set of analyses. (The basic limitation to the Mantel-Haenszel technique is that one must explicitly create a 2 x 2 table at each level of the confounding variable, or combina­tion of confounding variables.

With five binary variables, at least 32 tables are required, and if the data set contains only a few hundred sampling units —groups of calves in the case of the Bruce County Study —many of the cell entries will be zero. Logistic regression, in a manner similar to multiple least squares regression, allows one to obviate this problem.)

Detailed discussions of the results of the above project are available and are not germane to the objectives here. The major point to stress is that formal analyses at the group and/or farm level are extremely useful in providing information for rational decision making. However, no one study should be viewed in isolation.

Results from all studies, be they obser­vational, experimental, or theoretical, must be integrated with local expe­rience and interpreted in combination. (Throughout this text, constraints have been mentioned in terms of one’s ability to learn by experience. While it is true for manual skills that practice makes perfect, the same is not necessarily true when making management decisions. Although experience ought not be ignored, one needs to recognize its tendency to lead to authoritarian rather than authoritative discussions.)

During the past decade, a number of well-designed farm-level studies of dairy farms have been initiated or reported. If these studies have a drawback, it is that the number of herds involved was too few to allow formal analyses of factors that might have impacted on productivity or disease occurrence.

Nonetheless, there is an excellent series of reports on the Australian experience with planned animal health programs. Recently, two reports on herd-level studies in Minnesota dairy herds have also been published.

Investigations into calf survival have also been conducted at the herd level, although not many studies have formally analyzed differences in mor­bidity and mortality among herds for their relationship to management practices. Nonetheless, insight into how to conduct field studies of calf survival and the problems associated with them can be found in recent articles.

A study of calf survival in Norway utilized data from a large number of herds; however, the emphasis appeared to be on individual calf survival and factors relating to this (the outcome was lived or died for each calf in the study). Herd-level and individual animal factors were used as predictor variables but did not appear to be important.

The results of a recent study in Ohio suggest that management factors are more predictive of disease problems in calves than is the presence or ab­sence of putative pathogens. Again, this was difficult to formally assess because of the small number of herds in the study.

Currently, a study of calf survival and factors influencing it is being conducted on 104 dairy farms in Ontario as part of a larger overall dairy farm study referred to previously. At the beginning of the study, each farm was visited and a calf management policy question­naire was administered by personal interview.

At that time, the physical calf rearing facilities were also evaluated. At the end of the first year, each farmer was mailed a “re-check” questionnaire containing a subset of ques­tions from the original survey. At the end of the second year, all farms were visited and, where possible, fecal samples from the youngest one or two calves under 2 weeks of age were obtained for microbiologic screening.

These samples were used to assess the relationships between pathogen sta­tus and disease. All farmers kept daily log sheets of all calf births, preven­tive and disease treatments, and deaths among pre-weaned calves, and these sheets were picked up during regular visits by the project field technicians. At the end of the survey, as part of a more general management question­naire, the dairy farmers were asked to note any recently implemented calf management policy changes.

It is anticipated that the results of this study will provide solid, scientifically valid evidence on the effect of a number of factors that are thought to impact on calf morbidity and mortality. A herd-level field trial of rota-corona virus vaccine and E. coli bacteria was conducted as part of this study.

Although this section has emphasized bovine health management, the philosophy of health management at the herd level is perhaps more ad­vanced in the swine industry, and examples of this will be presented in subsequent sections.

Also, despite the overwhelming emphasis on and im­portance of the individual in companion animal medicine, there is a great need for the formal application of epidemiologic methods in this area. Studies dealing with such items as population disease control, population control, animal behavior, and the human-animal bond are desperately needed.

Problem Resolution in Intensively Managed Units:

Although disease outbreaks still occur, many of the diseases that have high case fatality rates, or pose a significant direct public health threat, or interfere with international trade have been brought under control in many countries. If these diseases still exist, they often do so at hypo-endemic or sporadic levels.

Since 1960, it has become apparent that endemic, often subclinical, diseases have a large impact on the productivity of intensively reared animals. As mentioned, control of many of the epidemic diseases allowed a fundamental change in the structure of agriculture toward larger mono-species farms.

Thus, in the past few decades, veterinarians have be­gun to turn their attention toward the farm or flock as the unit of concern rather than the individual animal. This trend is particularly advanced in the poultry industry, commercial swine operations, and the beef feedlot in­dustry. Even in the dairy industry, where individual purebred animals still have great economic value, the trend is away from the individual toward the herd.

As part of this change in emphasis, veterinarians must acquire new skills to identify and deal with problems at the herd level; an extrapola­tion of skills appropriate to individual animals is not a satisfactory solu­tion. Basic epidemiologic training can provide many of these skills, but veterinarians will have to modify and extend many of the current problem- solving techniques of epidemiology to make them more suitable for use in intensive animal industries.

Today, there is only sparse information on the concepts and techniques of problem solving at the herd level in veterinary medicine. The following discussion should prove useful as an initial meth­odology in this regard, and it is hoped, will provide the stimulus for the required new developments in this area.

The discussion assumes that an adequate on-farm data recording and analysis system exists, because in the absence of such a system problem solving at the herd level becomes a difficult, often hit-and-miss operation. The record system need not be computerized, but it is likely most farms will utilize a computerized system in the future.

The development of both computer software and hardware products appropriate to veterinarians and their clients is an active and evolving area. It is not the intent to describe or evaluate these systems here, but rather to provide a sound basis for their introduction, adaptation, and usage.

The evolution of one major system (DAISY) designed for the dairy industry is a useful study for those contemplating work in this area, Programs for the swine industry are also appearing rapidly, particularly after descriptions of the design and use of a breeding records system in England were published. A recent comprehensive overview of swine recording systems in the United Kingdom and a formal evaluation of a number of dairy recording systems are also available.

A schematic outline of the steps involved in designing and using a health- oriented data base is shown in Figure 12.3. These include formulating a set of written production-based objectives, deciding on critical levels for a number of parameters that signal the need for investigation, preparing action lists to remind the client and the veterinarian of routine duties as well as identifying problem areas and/or problem animals, and monitoring the production response.

If current objectives are not being met, the herd management and/or health maintenance program will require modifica­tion. If the current objectives are being met, steps may be required to safeguard the herd; in other cases production targets may be raised.

Two important features of a health management strategy are:

First, it is unlikely that by helping to achieve production goals the veterinarian will have no work. Rather, most clients will ask the veterinarian to remain as an integral component of the management of the production unit.

Second, it is of paramount importance that the veterinarian and client learn from their activities, be they successes or failures. Otherwise there is a tendency to redouble efforts yet go nowhere, as if on a treadmill.

If problems exist, be they production deficits or increased disease occur­rence, an outline of procedures to resolve them is presented in Figures 12.4 and 12.5. In this outline it is assumed that a dairy herd is the unit of concern; however, similar charts could be drawn by analogy for poultry, swine, and beef units.

The first step in problem resolution is to identify that a problem exists and to define in general terms what the problem is. In this regard, a few production parameters that are both biologically and economically mean­ingful should be monitored on a regular basis.

For a dairy herd, monitoring suitable herd production parameters (such as milk production’ per unit time and survivorship in adult animals) will indicate when a problem exists. (For calves, growth rates and survivorship would be appropriate parameters to measure.)

Milk production per cow per day is probably the most useful overall measure of productivity, biologically and economically, because it incorporates measures of milk production and reproductive performance. If this is low, one would then proceed to identify whether the major problem lies in reproductive performance, milk production, or both.

The temporal pattern of milk/cow/day can easily be monitored by dividing the volume of milk shipped each day by the number of cows in the herd and plotting the result against time. The resultant graph can quickly identify sudden changes in productivity (e.g., reduced milk production), and it can also be used to monitor long-term trends (such as a gradual reduction in productivity due to declining reproductive performance).

Once it is known that a problem exists, the second step is to examine additional parameters to determine what the problem is. For example, the herd average calving-to-conception interval or the percent of the herd preg­nant by 120 days are useful parameters for assessing the overall efficiency of a dairy herd’s reproductive program. While, the average calving-to-conception interval is perhaps the easier parameter to interpret, determination of which cows to include in the calculation can be difficult.

Cows that never conceive will not be included in the calculation, and consequently the pa­rameter may overstate the true efficiency of the breeding program. On the other hand, percent of the herd pregnant by 120 days (or any other agreed upon cut-off point) circumvents this problem and identifies a production deficit quickly.

However, it suffers from the drawback that a cow open 200 days has no greater impact on the parameter than a cow open 121 days. Additional parameters worth monitoring in a dairy herd include: milk/ cow/day (a measure of nutritional status and other general management factors); bulk tank somatic cell count or the herd geometric mean somatic cell count (indicators of subclinical mastitis); and the lactational incidence rates of the more common clinically evident diseases.

Although changes in any of the parameters described above will even­tually result in a change in the overall measure of productivity (i.e., milk/cow/day), there will inevitably be a delay before the change is apparent. Since most of the parameters are readily available, many producers will choose to monitor the more specific parameters on a regular basis.

For example, an increase in subclinical mastitis will inevitably result in a reduc­tion in milk/cow/day. However, since many other changes may be taking place in a herd at the same time (e.g., cows drying off and freshening, ration changes, etc.), the reduction in milk/cow/day may not be evident for some time.

The bulk tank somatic cell count is a more sensitive indica­tor of the level of subclinical mastitis and will reflect the change more quickly. Consequently, there is merit in monitoring these more specific pa­rameters to prevent a drop in productivity.

The third step in problem resolution is to determine in very specific terms what the problem is and why it has occurred. An analogous situation in individual animal medicine would be progressing from an observation that a dog has a persistent ocular discharge to a diagnosis of kerato-con­junctivitis sica due to inadequate tear production. However, instead of using clinical examinations and diagnostic tests to refine the diagnosis, the veterinarian analyzes herd records and the results of screening tests.

To further define the problem on a herd basis, it is necessary to identify when and where the problem occurs and which animals are affected. In answering these questions, parameters called diagnostic indices are used to assess specific aspects of the production system. As an example, the first service conception rate in a dairy herd is a good indicator of fertility in cows presented for breeding.

The herd’s average values of these diagnostic indices should be compared to preset targets or goals. In addition, for production units with sufficient animals, it is useful to note the standard deviation of the indices. An abnormal average with an accept­able standard deviation indicates a general herd problem, as, for example, one that would result from inadequate nutrition or a herd-wide manage­ment problem.

A large standard deviation indicates that individual animals or a subset of the herd constitutes a major part of the problem, and one should identify these abnormal animals and try to determine reasons for their poor performance. A small standard deviation is as important to economic return and ease of management as meeting a stated production average.

For herd medicine problems, one of the most important determina­tions to be made is, when in the production cycle does the problem occur? For a reproductive problem, this question becomes, at what point between calving and eventual conception are events not occurring as expected? To answer this question, the calving-to-conception interval is subdivided and various parameters that assess specific portions of the reproductive pro­gram are calculated (Fig. 12.5).

For example, if a herd has a prolonged calving-to-conception interval (160 days) but an acceptable average number of days to first breeding (70 days), parameters such as number of services per conception, percentage of cows presented for pregnancy diagnosis that are found to be “open” (a measure of estrus detection in the herd), and incidence rates of cystic ovaries and other reproductive diseases should be examined.

While examining the question, when does the problem occur? it is also appropriate to examine the temporal distribution of the problem. This may involve determining long-term trends, seasonal variations, and even short-term variations.

For example, if a herd has an excessive number of services per conception, it would be appropriate to examine conception rates by day of the week. It is possible that the individuals responsible for inseminations on the weekend are not as skilled as their weekday counter­parts. Conception rates may also vary seasonally in response to changes in nutrition and housing.

Determining which animals are involved in a problem requires a crite­rion by which animals can be classed as “normal” (e.g., 80 days for calving to first breeding or 200,000 cells-/ml for a somatic cell count) or “abnor­mal”. Then the percentage of abnormal animals in various groups within the herd can be calculated.

It may be informative to compare animals of different age groups, different breeds, high producers versus low producers, etc. For example, determining the prevalence of elevated cell counts in cows in various age groups can be helpful in arriving at a “herd diagnosis”.

The question of where the problem is occurring can be answered in a similar manner. The relative frequency with which the problem appears in animals in different pens or barns, or different locations within a barn or a milking string should be determined. This information can then be studied and possible explanations for the pattern such as ventilation problems or inadequate water sources can be sought.

Answering the questions when, who, and where may not completely define the problem at hand and additional data may be required. Once collected, these additional data can be combined with the information about when, who, and where for a detailed specific definition of what the problem is. As an example, a veterinarian may start a problem-solving exercise with the observation that the calving-to-conception interval for a herd is too long and has a large standard deviation.

Through the analyses of appropriate records it may be possible to identify that the specific prob­lem relates to very low conception rates in cows bred on Friday, Saturday, and Sunday. The veterinarian and producer would then have to collect additional data or conduct a small trial to determine if the problem relates to cows being bred at the wrong time in their cycle on the weekends (i.e., a problem in estrus detection) or to inappropriate technique on the part of the inseminator.

Once a clear statement has been made as to what the problem is, the number of possible explanations as to why it is occurring will be greatly reduced. In the example above, if it turns out that cows not in heat are being bred on the weekend, the possible explanations might include inabil­ity of the person involved to correctly identify the signs of heat, or incorrect recording of cow names and numbers.

This approach to problem solving is not restricted to situations where dramatically serious deficits exist. As the following two examples show, it can be used to help rectify relatively minor or moderate problems to im­prove the productivity of the herd.

Problem Resolution: Example 1:

A 90-cow Holstein-Friesian herd in Ontario, Canada had a rolling herd average of 147 BCA units and a daily milk production of 20.7 liters/cow/day. Both production parameters indi­cated a reasonable level of milk production, but the bulk tank somatic cell count (SSC) had averaged 509,000 cells/ml over the last 6 months.

The producer was not particularly concerned, but the veterinarian pointed out that with milk valued at $40/hL, there was a loss in excess of $8000/yr in milk compared to production at a cell count average of 150,000 cells/ml.

To further investigate the problem, the veterinarian classified all the cows as having “elevated SCC” if their most recent individual cow somatic cell count was over 200,000 cells/ml and “normal” if it was less.

While investigating when the problem occurred, it was found that the distribution of counts according to the cows’ stage of lactation was as follows:

In general, counts were low early in lactation, suggesting the dry cow therapy program on this farm was adequate and also that the majority of new infections were not occurring around the time of parturition. The dramatic rise in the prevalence of elevated counts throughout the lactation is suggestive of cow-to-cow transmission of a pathogenic agent.

To determine which cows were infected, the cows were classified ac­cording to age and cell count status with the following results:

It was quite evident that the prevalence of elevated counts increased with the age of the cows, but since one expects very few elevated counts in first calf heifers, the 25% prevalence observed in this herd was additional cause for concern.

At this point it was concluded that the herd had a high prevalence of infection, with cow-to-cow spread during the lactation being the most likely mechanism of transmission. It was also concluded that most infections were eliminated by the dry cow therapy and that management at the time of calving was adequate since relatively few new infections occurred then.

To further characterize the problem, data about the incidence of clini­cal mastitis were collected, and composite milk samples from the 75 milk­ing cows were collected for culturing. The incidence of clinical mastitis was 2.7% per month (i.e., 2.7 cases/100 cows/mo), which was deemed accept­able. Of the 31 samples that were culture positive, 26 (84%) yielded Strep­tococcus agalactiae.

The “herd diagnosis” of this problem could now be stated as a high rate of cow-to-cow transmission of S. agalactiae during lactation, resulting in a high prevalence of subclinical mastitis with an attendant economic loss in excess of $8000/yr. Resolution of the problem depended on identifying those faults in the milking system and the operator’s technique that related either to cow-to-cow transmission of the organism or to increasing the susceptibility of the cows to new infections.

Problem Resolution: Example 2:

A 200-cow Holstein-Friesian herd in Ontario, Canada had a calving-to-conception interval of 121 days. The dairy farmer, in conjunction with the veterinarian, had set 90 days as the herd objective and, based on an estimated loss of $2.50/cow/extra day open, it was estimated that suboptimal reproductive per­formance was resulting in a loss of approximately $15,000/yr.

To identify when in the sequence of reproductive events the problem was occurring, the veterinarian examined several diagnostic indices:

From these data it was apparent that of the 31 days being lost, all of the loss was occurring prior to the first breeding. A proportion of this loss appeared to occur because of the delay between first heat and first breed­ing, but the greatest loss was due to failure to detect heat early in all cows. The standard deviations for both the calving to first heat and calving to first breeding intervals were too large (in excess of 30% of mean), indicat­ing considerable variability among cows within the herd.

In further investigating the loss of time between first heat and first breeding, it was recognized that since the producer had decided that cows would not be bred prior to 50 days postpartum, not all cows could be bred on their first detected heat. However, of the 210 cows calving, 26 (12.4%) had heats detected on or after day 50, at which time they were not bred.

An average of 47 days, about two estrus cycles (called “deferral days”), then elapsed before those cows were again detected in heat and bred. The total days lost by this failure to breed at the first appropriate heat in this small group of cows resulted in an extra 6 days in the calving-to-conception interval when averaged over the whole herd.

Not satisfied with simply identifying one source of inefficiency in the reproductive program, the veterinarian further investigated the problem of deferral days. When the cows were subdivided into heifers and mature cows it was found that 0% and 18.5% of each group, respectively, were deferred. Heifers have a greater persistency of milk production than do cows; conse­quently a longer calving-to-conception interval in heifers has less detrimen­tal effect on overall productivity.

Thus, the veterinarian was concerned that the deferrals were occurring in mature cows instead of in heifers. However, the veterinarian had been continually stressing the importance of early breeding, and when the percentage of cows deferred during the first 6, second 6, and last 4 months of the study period were calculated, the results were 17.0%, 12.8%, and 0%. It appeared that the problem of deferrals had been solved. The veterinarian then turned to the problem of identifying why cows were not being seen in heat early enough.

A number of factors were ex­amined and the results of several were as follows:

Age did not appear to be a factor in the problem. However, the 26 cows that had retained placentas had a substantially longer interval to first ob­served heat, suggesting that measures to reduce the incidence of retained placenta might be in order. The problem was also more serious in the higher producing cows, suggesting that the nutrition program in the dry period and early lactation should be reviewed.

Finally, the problem appeared to be more serious in the winter. The veterinarian had noticed that the operators were less likely to be around the barn later in the evening during the winter and one possible consequence of this was a reduced level of heat detection.

These analyses were not a complete evaluation of all aspects of the reproduction program on the farm, but they did serve to identify the major problem areas.

The problem of deferral days was identified, and with as­sistance it appeared that the producer had rectified that situation. It was also determined that cows having a retained placenta and cows calving during the winter were more likely to have a prolonged calving-to-first-heat interval. Steps to rectify those problems could be initiated immediately.

Finally, the problem of failure to detect heats appeared more serious in high producing cows. A review of the nutritional program along with an evalua­tion of body condition scores would be required before corrective measures for that problem could be undertaken.

Once the problem area(s) have been identified, corrective action must be taken (step 4, Fig. 12.4). To institute the appropriate directed action, the practitioner’s current knowledge may suffice, or the assistance of other personnel may be required. In some cases, the control strategies will not be obvious and further study of the type exemplified in the previous section will be required.

Multiphasic screening (biochemical-metabolic profiles) and serologic data can be combined to allow the simultaneous study of the physiologic status and the infection (immune) status of individuals and the herd. Questions such as how many animals to sample, which animals to sample, and how many samples per animal are required for this purpose, remain largely unanswered.

Nonetheless, first approximations are possible using the sampling techniques. The fact that multiphasic screening generally has failed to produce obvious benefits may in part reflect historic limitations with regard to sampling, testing , analy­sis, and interpretation of results, rather than the true value of the proce­dure.

If no answer to the problem is obvious, practitioners should be pre­pared to conduct well-designed, analytic observational studies and/or field trials. It is quite likely that in the future farmers will not demand immediate answers of the veterinarian, but they will demand that the veterinarian know how to find the answers.

If the problem is at the herd level, the veterinarian should be able to obtain assistance from personnel at an epide­miologic research unit. As mentioned, a major reason for the existence of such a unit is to assist in problem solving at the herd level.

The latter is very difficult for an individual practitioner to perform because data on many herds (flocks) are required, and the analytic expertise and com­puter requirements to manipulate and analyze the large volume of data or the ability to conduct multi-herd field trials will be beyond the capabilities of most practitioners.

The final stage (step 5, Fig. 12.4) of problem resolution is to monitor the progress of the herd to ensure that the corrective easures have had the desired effect. If production levels fail to increase (or if production is not more efficient), the practitioner should re-examine the diagnostic indices to ensure that the correct problems have been identified. If this is confirmed, the control measures should be re-examined and alternate strategies em­ployed if deemed necessary.

It is highly likely that mistakes will be made as veterinarians enter this new era of health management. Some mistakes are inevitable. The key is that the individual veterinarian, the client, and the veterinary profession must learn from their experiences, so clients get the best current informa­tion and advice and the quality of information improves with time.

It should be obvious from the preceding discussion that the practitioner of the future is an applied researcher as well as a provider of essential technical services and information. Indeed, the combination of these two activities will likely increase the satisfaction of practitioners and prolong their pro­ductive days in practice.