SciELO - Scientific Electronic Library Online

vol.51 issue2Thyroid epithelium height as a sorting method for detection of antithyroid residues in bovinePrevalence and classification of foot problems in lactating dairy cows in Belo Horizonte, Brazil author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Arquivo Brasileiro de Medicina Veterinária e Zootecnia

Print version ISSN 0102-0935On-line version ISSN 1678-4162

Arq. Bras. Med. Vet. Zootec. vol.51 no.2 Belo Horizonte Apr. 1999 

Mortality on grower/finisher-only swine operations in the United States

(Mortalidade em granjas de crescimento/terminação de suínos nos Estados Unidos)


W.C. Losinger, E.J. Bush, M.A. Smith, B.A. Corso

United States Department of Agriculture, Animal and Plant Health Inspection Service
Veterinary Services, Centers for Epidemiology and Animal Health
555 South Howes Street, Suite 200
Fort Collins, Colorado 80521 USA


Recebido para publicação em 26 de agosto de 1998.




For 53 grower/finisher-only swine operations that participated in the United States National Animal Health Monitoring System 1995 National Swine Study, mortality among finisher pigs ranged from 0 to 12.0% over a 6-month period. Twenty-six (49.1%) had <2% mortality, and 27 (50.9%) had >2% mortality. Nine (17.0%) operations experienced >4% mortality. Fisher's exact test revealed that operations with all-in all-out management were significantly more likely to have <2% mortality than operations with continuous management, and that operations where all finisher pigs came from farrowing units belonging to the operation (either on-site or off-site) were significantly more likely to have <2% mortality than operations where >1 grower/finisher pig came from another source. Larger operations (where >900 pigs entered the grower/finisher phase) practiced all-in, all-out management more frequently than smaller operations, and had a lower mean percent mortality than smaller operations. Diagnosis of Salmonella in finisher pigs performed at a laboratory or by a veterinarian in the 12 months prior to interview was associated with both increased percent mortality and increased percent mortality per day.

Keywords: Pig, swine grower/finisher, mortality, animal health



A mortalidade de suínos, durante um período de 6 meses, variou de 0 a 12% em 53 granjas de crescimento/terminação, que participam do "United States National Animal Health Monitoring System 1995 National Swine Study". Vinte e seis (49,1%) granjas tiveram <2% de mortalidade e 27 (50,9%) tiveram índices >2%. Nove (17,0%) granjas apresentaram mortalidade >4%. A análise pelo teste de Fisher revelou que as granjas que adotavam o sistema "all-in, all-out" eram significativamente mais propensas a apresentar índices <2% que as granjas com manejo contínuo, e que as granjas nas quais todos os suínos terminados eram oriundos de unidades pertencentes às próprias granjas eram significativamente mais propensas a ter mortalidade <2% que as granjas onde os animais (>1) provinham de outras origens. Granjas maiores (com mais de 900 animais entrando na fase de crescimento/terminação) adotavam com maior freqüência o manejo "all-in, all-out e tiveram menor percentual de mortalidade do que pequenas granjas. O diagnóstico de Salmonella em suínos terminados, realizado 12 meses antes da intervenção, foi associado ao aumento do percentual de mortalidade geral e diário.

Palavras-Chave: Suínos, crescimento/terminação, mortalidade, saúde animal




Pork production in the United States has evolved from being largely a complementary portion of integrated-crop/livestock farms to the point where the majority of pigs are raised on large, intensive operations (Honeyman, 1996). Some swine operations now specialize in one phase of production only (for example, producers of breeding stock, producers of feeder pigs, grower/finisher-only, etc.). Operations that specialize in one phase of production may be managed differently from other types of swine operations, and therefore may encounter unique sets of problems.

The 1995 National Swine Study, which was completed under the auspices of the National Animal Health Monitoring System (NAHMS) of the United States Department of Agriculture (USDA), was designed both to profile management practices employed on swine operations throughout the major pork-producing states (Swine..., 1995) and to provide information specifically on the grower/finisher- production phase on operations with >300 finisher pigs (Reference..., 1996). Losinger et al. (1997 and 1998) used data from the 1995 National Swine Study to analyze swine mortality during the grower/finisher production phase overall and on farrow-to-finish operations. The purpose of this paper is to report results of analyses of swine mortality specifically on grower/finisher-only operations. The results of these analyses may be beneficial to both researchers and pork producers in identifying areas for further attention and scientific investigations.



During the first stage of data collection for the 1995 National Swine Study (1-16 June, 1995), USDA:National Agricultural Statistics Service (NASS) enumerators collected data on general management practices employed on operations with pigs in 16 major swine-producing states (Swine..., 1995). Participation in the second stage of data collection, which concentrated on the grower/finisher phase of production, was limited to first-stage participants with >300 finisher pigs (with at least one finisher pig >54.4kg). The second stage of data collection involved on-farm interviews by federal and state veterinary-medical officers (VMOs) and animal-health technicians (AHTs): once between July and September, 1995 (to collect data on feed and waste management practices), and again between November, 1995 and January, 1996 (to gather more detailed information on quality control, swine health, feed management and marketing practices). During the final VMO or AHT interviews, producers provided the number of pigs that had entered the grower/finisher production unit and the number of finisher pigs that had died during the 6-month period prior to the interview.

Analyses were limited to operations that had characterized themselves as being primarily grower/finisher only during the NASS interview. For each operation, percent mortality was computed by dividing the number of finisher pigs that had died during the 6-month period prior to the final interview by the number of pigs that had entered the grower/finisher phase of production, and multiplying by 100. A mortality rate (percent mortality per day in the grower/finisher unit) was computed by dividing percent mortality by average days in the grower/finisher unit (which was asked on the NASS questionnaire). Responses to other questions from the questionnaires were turned into 98 dichotomous variables. Fisher's exact test, provided by the FREQ procedure of SAS (SAS..., 1990), was used to compare the percent of operations achieving <2% mortality for each of the dichotomous variables.

Dichotomous variables with P<0.30 (based on Fisher's exact test) and that had >10 responses for each category were considered for inclusion in multivariable models. Spearman rank correlation coefficients (Hogg & Craig, 1978), computed from the CORR procedure of SAS... (1990), were used to examine relationships between these variables. The REG procedure of SAS... (1989) was used to perform stepwise regression twice. Percent mortality served as the dependent variable in one model. Mortality rate (percent mortality per day) served as the dependent variable in the other model. Dichotomized variables, representing region and operation size (as determined by the number of pigs that entered the grower/finisher unit in the 6 months prior to interview), were forced into the models to make certain that other variables did not enter the models merely because of regional or operation-size differences in management. The significance level required for other explanatory variables to enter and remain in the models was 0.05. The GLM procedure of SAS... (1989) was used to generate least squares means estimates of percent mortality and percent mortality per day for the variables included in the models.



Fifty-nine (14.1%) of the 418 operations that took part in the second stage of data collection for the 1995 National Swine Study categorized themselves as grower/finisher only (Reference..., 1996). Of these, 53 (89.8%) furnished mortality data for the 6-month period prior to the final interview. Operation mortality ranged from 0 to 12.0%, with a mean of 2.8% and a standard deviation of 2.5%. Table 1 summarizes the percent of grower/finisher-only operations by percent mortality. Twenty-six (49.1%) had <2%, and 27 (50.9%) had >2% mortality. Days in the grower/finisher phase ranged from 50 to 195 days, with a mean of 111.1 and a standard deviation of 26.7. Mortality rate ranged from 0 to 0.1280 percent mortality per day in the grower/finisher unit, with a mean of 0.0258 and a standard deviation of 0.0225.



Table 2 gives the number and percent of operations with <2 % mortality and the number of operations with >2% mortality for the dichotomous variables with P<0.30 and >10 responses in each category. A significantly higher proportion of operations with all-in, all-out management had <2% mortality than operations with continuous management (68.0% compared to 32.1%). A significantly higher percent of operations where all finisher pigs came from farrowing units belonging to the operation (either on-site or off-site) achieved <2% mortality than operations where at least some pigs came from other sources (73.3% versus 39.5%). No statistically significant differences were detected in any of the other variables.



Table 3 presents the least-squares means of percent mortality for variables included in the regression model. Although operations with all-in, all-out production were more likely to achieve <2% mortality than operations with continuous management, this variable was correlated with both region (r=0.27) and particularly herd size (r=0.44) (both of which were forced into the multivariable model), and did not enter the multivariable model. In this study, 6 of 20 (30.0%) grower/finisher-only operations in the North/Southeast region reported all-in, all-out management, compared to 19 of 33 (57.6%) in the Midwest (P=0.088). Six of 25 (24.0%) operations where <900 pigs entered the grower/finisher phase of production during the 6-month period reported using all-in, all-out management, were compared to 19 of 28 (76.0%) operations where >900 pigs entered the grower/finisher phase (P<0.001). In addition, operations where >900 pigs entered the grower/finisher phase had lower average percent mortality than operations where <900 pigs entered the grower/finisher phase (Table 3). Diagnosis of Salmonella in the 12 month-period prior to the final interview was the only non-forced variable to enter the model. The model with percent mortality per day as the dependent variable (Table 4) was very similar.





Because this analysis was limited to grower/finisher-only operations with >300 finisher pigs, the results are not necessarily germane to other types of swine operations. Similarly, the results do not necessarily apply outside the 16 states included in the 1995 National Swine Study.

To facilitate analysis, all explanatory variables were turned into dichotomous variables. Even then, only two variables were found to have a statistically significant association with <2% mortality. Since the sample size is related to the significance level of statistical tests (Snedecor & Cochran, 1967), more variables would probably have been found to be significantly related to <2% mortality if a larger number of grower/finisher-only operations had participated in the study.

Biosecurity measures, including being careful about where one obtains pigs, are important in controlling the introduction of diseases to swine operations (Robertson, et al. 1992). Although one might have expected that operations that considered themselves primarily grower/finisher would have obtained their pigs principally from farrowing units that did not belong to the operation, 15 (28.3%) reported that all of their pigs came from farrowing units that they owned (either on or off the operation). Operations that receive pigs from elsewhere risk introducing infectious agents into their herds. Clermont & Désilets (1982) reported increased mortality on operations where pigs had been bought at auction sales or from livestock dealers (as compared to operations that did not obtain pigs from these types of sources). The results of the present study indicated that grower/finisher-only operations that obtained all of their pigs from farrowing units belonging to the operation were more likely to attain <2% mortality than operations that obtained some finisher pigs from other sources (such as feeder-pig producers, auctions, sale barns, livestock markets, and other sources). However, the sample size was not sufficient to build a multivariable model (with a binomial dependent variable) to test whether obtaining all pigs from farrowing units belonging to the operation was genuinely associated with <2% mortality when other explanatory variables were considered.

Désilets & Clermont (1982) reported reduced mortality among finisher pigs under all-in, all-out management as compared to continuous management. All-in, all-out production during the grower/finisher phase is generally considered to be an effective means of reducing disease, improving growth, and diminishing mortality (Scheidt, et al., 1995). Since larger operations in the present analysis were more likely to practice all-in, all-out management than smaller operations, it is possible that the lower percent mortality observed for larger operations could have been influenced by their penchant for all-in, all-out management. It is also possible that other differences in management could have accounted for the observed difference in percent mortality by operation size. When mortality was adjusted for length of time in the grower/finisher unit (by dividing percent mortality by number of days in the grower/finisher unit), the operation-size difference in mortality rate was not significant (Table 4). Clermont and Désilets (1982) found the mortality rate to be directly proportional to herd size.

Salmonella is an important pathogen to pigs in the United States (Kramer, 1995). Wilcock (1986) noted that most outbreaks of porcine salmonellosis occurred in intensively-reared weaned pigs. The results of the 1995 National Swine Study revealed that Salmonella had been diagnosed in finisher pigs by a veterinarian or laboratory during 1995 in 9.6+2.5% of operations in the targeted population (USDA, 1996). Ten (18.9%) of the grower/finisher-only producers in this study said that Salmonella had been diagnosed in finisher pigs on their operation (Table 2). Salmonella is very hardy and ubiquitous in nature, and, according to Wilcock (1986), prevention of Salmonella infection in pigs is virtually impossible. Some operations in the study probably had Salmonella without a veterinarian or laboratory diagnosis. Information on the severity of the infection on the operation was not collected in the study. Some of the operations that reported diagnosis of Salmonella may have been experiencing rather severe conditions to warrant a diagnosis by a veterinarian or laboratory. It is possible either that conditions that led to the diagnosis of Salmonella in an operation also led to increased mortality, or that Salmonella was a direct cause of mortality. Porcine salmonellosis is primarily a septicemic disease caused by Salmonella choleraesuis, and is common particularly in the Midwestern states (Kramer, 1995). Wilcock (1986) reported that mortality was high in most outbreaks of septicemic salmonellosis.



CLERMOT, R., DÉSILETS, A. Epizootic aspects of swine respiratory infections which occurred in Quebec from september, 1980 through february, 1981. Rev. Vet. Can., v.23, p.179-182, 1982.        [ Links ]

DÉSILETS, A., CLERMONT, R. Influence of several extrinsic factors on mortality rates in finisher pigs. Med. Vet. Que., v.12, p.118-120, 1982.        [ Links ]

HOGG, R.V., CRAIG, A.T. Introduction to mathematical statistics. 4.ed. New York: Macmillan, 1978.        [ Links ]

HONEYMAN, M.S. Sustainability issues of U.S. swine production. J. Anim. Sci., v.74, p.1410-1417, 1996.        [ Links ]

KRAMER, T.T. Swine salmonellosis: what is new about it? Agri-practice, v.16, p.13-16, 1995.        [ Links ]

LOSINGER, W.C., BUSH, E.J., SMITH, M.A., et al. An analysis of mortality in the grower/finisher phase of swine production in the United States. Prev. Vet. Med., v.33, p.121-135, 1997a.        [ Links ]

LOSINGER, W.C., BUSH, E.J., SMITH, M.A., et al. Mortality in the grower/finisher phase on farrow-to-finish swine operations in the United States. Ann. Zootech., v.47, p.99-105, 1997b.        [ Links ]

REFERENCE of 1995 U.S. grower/finisher health & management practices - N201.696. Fort Collins: USDA/APHIS/VS/CEAH, 1996.        [ Links ]

ROBERTSON, I.D., MHOMA, J.R.L., HAMPSON, D.J. Risk factors associated with the occurrence of swine dysentery in Western Australia: results of a postal survey. Aust. Vet. J., v.69, p.92-93, 1992.        [ Links ]

SAS procedures guide. 6.ed. Cary: SAS, 1990.        [ Links ]

SAS/STAT user's guide: version 6. ed. Cary: SAS, 1989. V.2.        [ Links ]

SCHEIDT, A.B., CLINE, T.R., CLARK, L.K. et al. The effect of all-in-all-out growing-finishing on the health of pigs. Swine Health Prod., v.3, p.202-205, 1995.        [ Links ]

SNEDECOR, G.W., COCHRAN, W.G. Statistical methods. 6.ed. Ames: Iowa University, 1967.        [ Links ]

SWINE 95. Part I: Reference of 1995 swine management practices - N186.995. Fort Collins: USDA-APHIS-VS/CEAH, 1995.        [ Links ]

WILCOCK, B.P. Salmonellosis. In: LEMAN, A.D., STRAW, R.D., GLOCK, W.L. et al. (Eds). Diseases of swine. 6.ed. Ames: Iowa State University, 1986.        [ Links ]

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License