Clinical factors, service type, and clinical outcomes associated with canine parvovirus in dogs presented at the KNUST Veterinary Teaching Hospital in Kumasi, Ghana Authors: Jemima Dzigbordi Agbota, Raphael Deladem Folitse, Emmanuel Darko Opoku, Dorcas Oyueley Kodie, William Tasiame, Benjamin Obukowho Emikpe Published: 2026 ABSTRACT BackgroundCanine parvovirus (CPV) continues to be a predominant cause of morbidity and mortality among dogs worldwide. In Ghana, data on clinical predictors, service type (inpatient or outpatient), and clinical outcomes in dogs afflicted by CPV is limited. This cross-sectional study, conducted at a Veterinary Teaching Hospital in Kumasi, Ghana, assessed the clinical factors associated with CPV diagnosis, service delivery types, and clinical outcomes in dogs. Diagnosis was based on compatible clinical signs and results from a Vcheck CPV antigen test kit with the V200 analyser (Bionote Inc., Korea), with dogs classified as CPV-positive or CPV-negative based on test results. Demographics, vaccination status, clinical signs, service type, and outcome data were recorded. Odds ratios were reported with 95% confidence intervals, and a significance threshold of p < 0.05 was set for regression analyses.ResultsA total of 129 CPV-positive dogs were diagnosed, with an overall mortality rate of 26.4% (34/129), compared to 8.5% (7/82) among CPV-negative dogs. Haemorrhagic diarrhoea (OR = 6.84), vomiting (OR = 2.75), lower body weight (OR = 0.94 per kg increase), and complete loss of appetite (OR = 4.29) were significant independent predictors of CPV infection. CPV-positive dogs were significantly more likely to require inpatient care than CPV-negative dogs (OR = 7.57, p < 0.001).ConclusionsCPV has a significant impact on dog morbidity and mortality in Ghana, with important emotional and financial consequences for affected dog owners. Classic gastrointestinal signs may assist in the clinical assessment of suspected cases in practice. CPV-positive dogs had higher inpatient care rates and mortality compared to CPV-negative dogs. These findings emphasise the need for improved vaccination uptake, client education, and clinical management practices to reduce CPV-related mortality in affected populations. Introduction In veterinary practice, dogs are not only patients but also integral to the emotional and social fabric of their owners’ lives, creating a unique veterinarian–client–patient bond that influences daily clinical interactions [6, 15]. The survival or loss of a canine patient often results in deep emotional weight for both families and clinicians, and infectious diseases that cause sudden, severe illness can strain this interconnected bond [31, 33, 38]. Among these diseases, canine parvovirus stands out as a particularly devastating threat. Its rapid progression, high mortality in untreated cases, and frequent presentation in young or cherished dogs make it one of the most distressing conditions for veterinarians to manage [21, 32, 34]. Canine parvovirus type 2 (CPV-2), a member of the genus Protoparvovirus within the family Parvoviridae, is a highly contagious virus that causes canine parvovirus (CPV) in dogs, with three major antigenic variants (CPV-2a, CPV-2b, and CPV-2c) circulating globally [8]. Its small, nonenveloped structure and stable protein capsid confers exceptional environmental stability, rendering it resistant to many disinfectants and facilitating persistent transmission in canine populations [39, 41]. Clinically, CPE infection manifests predominantly as an enteric syndrome characterised by vomiting, diarrhoea (mucoid to profuse haemorrhagic), anorexia, fever, abdominal pain, depression and, in severe cases, dehydration, electrolyte disturbances, sepsis, and endotoxemia [5]. In young puppies, especially those under three months of age, a less common but often fatal cardiac form may occur, causing sudden death or long-term pathological sequelae [13, 29]. The disease course is further influenced by host-related and socioeconomic factors, including age, breed predisposition, seasonality, and management conditions [18, 26, 43]. Mortality rates can reach as high as 91% in puppies affected without treatment [35], although outcomes improve significantly with timely intervention. Aggressive supportive care, including fluid therapy, antiemetics, antimicrobial therapy, and nutritional support, is typically used to manage infection, with reported mortality rates ranging from 4% to 53% [36]. Tertiary care hospitals have reported better survival rates for canine parvovirus (CPV) patients [18, 35]. However, the current gold standard treatment for CPV, which involves aggressive supportive therapy and hospitalisation, can be quite expensive, with costs ranging from $3000 to $5000 in high-income regions [34]. Compared with inpatient care, outpatient therapy, which is associated with increased mortality, can incur expenses ranging from approximately $300 to $400 for diagnosis and treatment [36]. These costs can be a significant financial burden for pet owners, with some even opting for euthanasia for their affected pets at times, highlighting the need to control the spread of the disease [19, 22]. Although published economic data on CPV management from African settings, including Ghana, remain scarce, treatment-associated costs may still constitute a considerable financial burden for affected pet owners in this region. Recently, monoclonal antibody therapy has shown promise in improving survival in experimentally infected CPV dogs at the very onset of viral shedding [21]; however, its effectiveness in naturally infected cases remains uncertain, as this therapeutic option is currently unavailable or unaffordable in many countries. CPV remains an important cause of canine morbidity and mortality in many African countries, including Ghana, where endemic circulation has been reported among hospital-presented dogs [12, 24, 40]. Despite its clinical importance, there is limited data on the clinical presentation, hospitalisation patterns and clinical outcomes of canine parvovirus infection in Ghana. In resource-limited settings, where access to advanced diagnostic tools, such as immunochromatographic assays or molecular tests, may be constrained, clinical features may be valuable for guiding diagnosis and management. This study was therefore conducted to determine the clinical factors and service type (inpatient/outpatient) associated with CPV, and to describe clinical outcomes among dogs presenting with suspected canine parvoviral infection at the KNUST Veterinary Teaching Hospital in Kumasi, Ghana. Materials and methods Study design and setting This was a cross-sectional study conducted from August 2023 to August 2024. It formed part of a broader observational study on CPV in dogs presented at the VTH, for which the prevalence and vaccination-related findings have been reported [2]. This study focused on the clinical factors, service type, and clinical outcomes of CPV patients diagnosed during the study period. This study was conducted at the Kwame Nkrumah University of Science and Technology (KNUST) Veterinary Teaching Hospital (VTH) in Kumasi, Ghana. The hospital functions as both a referral facility and a primary care facility for small animals, serving Kumasi and surrounding municipalities. Study population and case confirmation Dogs that presented to the KNUST Veterinary Teaching Hospital (VTH) during the study period with clinical signs suggestive of canine parvoviral enteritis, including vomiting, diarrhoea, inappetence, and lethargy, were considered eligible for testing. A total of 211 dogs that met the study eligibility criteria were included in this analysis. Diagnosis was confirmed using a Vcheck CPV antigen test kit, a rapid immunochromatographic assay, in conjunction with a V200 analyser (Bionote Inc., Korea), following the manufacturer’s instructions and protocols previously described by Agbota et al. [2]. The broader observational study from which this analysis was derived had an estimated sample size requirement of 365 dogs based on Cochran’s formula and a previously reported CPV prevalence; however, 211 dogs were eligible for inclusion during the study period [2]. Data collection For each enrolled dog, demographic, vaccination, and clinical information were recorded at presentation via a standardised data collection form, and outcome information was later added following treatment or follow-up. The data included age, sex, breed, vaccination status, clinical signs, type of service received (inpatient or outpatient), and clinical outcome (recovered, died, no treatment, or lost to follow-up). Inpatient and outpatient care definitions In this study, inpatient care referred to cases where dogs were admitted and remained in the hospital for continuous observation and treatment, typically receiving at least two treatment sessions per day. Unlike short-stay observation patients, these patients were not discharged after 12 h but remained hospitalised until they were clinically stable at discharge. Outpatient care referred to dogs that received a single in-clinic treatment session per day at the facility and were discharged home after each visit, with follow-up care such as oral antibiotics or rehydration solutions prescribed for administration by the owner at home under veterinary guidance. These dogs were returned on consecutive days for additional treatment sessions as required until recovery. The decision between inpatient and outpatient management was made jointly by the attending clinician and the owner on the basis of disease severity, owner preference, logistical factors, and financial considerations. Treatment protocol The treatment regimen was broadly standardised but adapted to individual case severity. Supportive care primarily consisted of fluid therapy, using intravenous Ringer’s lactate, dextrose saline (DNS), 0.9% normal saline, dextrose fluids (5% or 10%) or combinations thereof, adjusted according to hydration status and electrolyte requirements. The antibiotics used in treated cases or commonly prescribed during the study period included amoxicillin–clavulanic acid or penicillin–streptomycin, which are typically administered in combination with metronidazole, with ceftriaxone, ceftiofur or gentamicin reserved for selected cases. Antiemetic support was mainly provided with metoclopramide, whereas ondansetron was used when indicated. Adjunct therapies included antispasmodics (hyoscine butylbromide), gastric protectants (omeprazole), injectable high-potency vitamins (B-complex and multivitamins), oral rehydration solutions, and nutritional supplementation. Treatment intensity and frequency were greater in inpatients, who typically received multiple daily interventions, than in outpatients, who received single-session therapy with home care instructions. Clinical variables and definitions Clinical signs were assessed at presentation based on a combination of owner-reported history and clinician observation. Appetite status was categorised as normal, reduced appetite (hyporexia), or complete loss of appetite (anorexia). Reduced appetite was defined as decreased food intake compared to the dog’s usual feeding behaviour as reported by the owner, while complete loss of appetite referred to the absence of voluntary food intake. Diarrhoea was defined as the passage of loose or watery faeces. Diarrhoeic cases were further categorised as non-haemorrhagic diarrhoea or haemorrhagic diarrhoea based on the absence or presence of visible blood in the faeces, respectively, as reported by the owner or observed during clinical examination. Foul-smelling faeces were recorded based on subjective assessment by the owner and/or clinician, reflecting a noticeable deviation from the dog’s usual faecal odour. Vomiting was defined as the active expulsion of gastric contents and was recorded based on owner report and/or direct observation at presentation. Lethargy was defined as reduced activity level or decreased responsiveness compared to the dog’s normal behaviour, based on both owner history and clinical assessment, and was recorded as present or absent. Body weight was measured at initial case presentation using an electronic weighing scale, and rectal temperature was also recorded as a single measurement using a digital thermometer for this study. Outcome assessment Clinical outcomes were classified at discharge. Recovery was defined as the resolution of primary clinical signs, cessation of vomiting and diarrhoea, return of appetite, and rehydration. Death was recorded if the dog died during hospitalisation or if the owner reported death at home during the course of illness. Dogs that died before treatment initiation, were lost to follow-up, or received no treatment were also categorised. Dogs classified as having died before treatment initiation were those presented in a critical or moribund state and died prior to the commencement of therapeutic intervention. Dogs recorded as receiving no treatment were those for which the owner declined treatment after the initial evaluation, often due to financial or logistical constraints, and were not managed further at the facility. Cases classified as lost to follow-up were those for which outcome information could not be obtained after the initial presentation, despite attempts at follow-up through available contact details. Statistical analysis The data were entered into Microsoft Excel and analysed using IBM SPSS Statistics version 29.0 (IBM Corp., Armonk, NY, USA). Descriptive and inferential statistics were used to summarise demographic and clinical characteristics, service type, and outcome variables. Continuous variables (age, body weight, and body temperature) were assessed for normality using the Shapiro–Wilk and Kolmogorov–Smirnov tests, supported by visual inspection of histograms and Q–Q plots. Variables that were not normally distributed were summarised using medians and interquartile ranges (IQR), while normally distributed variables were presented as mean ± standard deviation (SD). For inferential analysis, logistic regression modelling was employed to evaluate factors associated with service type (inpatient versus outpatient) and CPV positivity. The analysis was conducted in two stages. First, univariable logistic regression was performed to screen for potential predictors. Variables with a p-value < 0.20 in the univariable analysis were retained for inclusion in the multivariable model to avoid premature exclusion of clinically relevant variables and potential confounders [16]. Categorical variables with sparse data across multiple levels, such as breed, were collapsed into broader clinically meaningful categories (local [indigenous] versus foreign [non-indigenous] breeds) to improve model stability and interpretability. Age in weeks was determined primarily from owner-reported dates of birth, available vaccination or health records, and clinician assessment. Age as a continuous variable was subsequently categorised into clinically relevant groups based on previously reported susceptibility patterns in CPV infection, severity, and outcome: ≤12 weeks, 13–24 weeks, and > 24 weeks of age [18, 28]. Multivariable logistic regression models were constructed using a backward stepwise elimination approach. Where clinically related variables were considered, alternative model specifications were explored, and the final model was selected based on statistical performance, clinical relevance, and overall interpretability. Model fit was assessed using the Hosmer–Lemeshow goodness-of-fit test and Nagelkerke’s R², while model discrimination was evaluated using receiver operating characteristic (ROC) curve analysis, with the area under the curve (AUC) reported. For the final model, multicollinearity was assessed using variance inflation factors (VIF) and condition indices, with all predictors demonstrating low collinearity (VIF: 1.055–1.164; maximum condition index: 10.545). Adjusted odds ratios (ORs) with 95% confidence intervals (CIs) were reported. A p-value of < 0.05 was considered statistically significant for all inferential analyses. Results Study population characteristics A total of 211 dogs presenting with clinical signs suggestive of canine parvoviral (CPV) infection were included in the study. Of these, 129 tested positive, and 82 tested negative for CPV. The overall study population comprised 131 male and 80 female dogs. The age of dogs was not normally distributed (Shapiro–Wilk testing and Kolmogorov-Smirnov p < 0.001) and ranged from 3 to 144 weeks, with a median age of 16 weeks (IQR: 10–24 weeks). Body weight was also not normally distributed (Shapiro–Wilk testing and Kolmogorov-Smirnov p < 0.001), with a median of 7.0 kg (IQR: 4.6–10.8 kg). In contrast, body temperature approximated a normal distribution (Shapiro–Wilk [p = 0.419] and Kolmogorov-Smirnov [p = 0.051] and is presented as mean ± SD (38.97 ± 0.81 °C). Dogs were distributed across age categories as follows: ≤12 weeks (n = 85), 13–24 weeks (n = 85), and > 24 weeks (n = 41). The most commonly represented breeds included local breeds (indigenous breeds) (n = 57), crossbreeds (n = 31), Boerboel (n = 25), German Shepherds (n = 21), Bullmastiff (n = 20), and Rottweiler (n = 16). Other less frequently represented breeds included Caucasian Shepherds (n = 13), Poodles (n = 12), Maltese (n = 4), Dobermanns (n = 2), Alsatian Shepherds (n = 2), American Bullies (n = 2), Belgian Malinois (n = 2), and isolated cases of Pugs, Shih Tzus, Siberian Huskies, and Great Danes (n = 1 each). Detailed demographic CPV risk factor analyses, including breed-specific distributions, have been reported previously [2]. Vaccination records showed that 102/211 (48.3%) dogs had documented prior CPV vaccination, while 90/211 (42.7%) were reportedly unvaccinated, and vaccination records were unavailable for 19/211 (9.0%) dogs. Detailed variations in vaccine doses received, vaccination compliance and vaccine timing were reported by Agbota et al. [2]. Regarding clinical management, dogs were managed as either inpatients or outpatients. Service type classification was available for 206 dogs that received treatment. Dogs that did not receive treatment or died before treatment were excluded from the service type analysis (Table 1). Table 1Distribution of service type among treated dogs (n = 206)VariableN (%) Inpatients 118 (57.3) Outpatients 88 (42.7) Total 206(100%) Only dogs that received treatment were included in the service type analysis; dogs that did not receive treatment or died before treatment were excluded Table 1Distribution of service type among treated dogs (n = 206)VariableN (%) Inpatients 118 (57.3) Outpatients 88 (42.7) Total 206(100%) Only dogs that received treatment were included in the service type analysis; dogs that did not receive treatment or died before treatment were excluded Distribution of service type among treated dogs (n = 206) Only dogs that received treatment were included in the service type analysis; dogs that did not receive treatment or died before treatment were excluded To further characterise clinical management patterns among CPV-positive cases, the distribution of service type across breeds was examined among treated CPV-positive dogs (Table 2). This informed the categorisation of breed variables for subsequent regression modelling, particularly given the distribution of cases across individual breed groups. Inpatient care predominated across nearly all breeds, accounting for 75.8% (94/124) of cases, while outpatient management was less frequent 24.2% (30/124). All cases in certain breeds, including Bullmastiff, Caucasian Shepherd, Maltese, Shih Tzu, Alsatian Shepherd, and American Bully, were managed as inpatients. The proportion of outpatient care was highest among local breeds (40.0%) and crossbreeds (30.0%). Table 2Descriptive distribution of CPV-positive dogs by service type and breed (n = 124)BreedInpatient N (%)Outpatient N (%)TotalLocal Breed22 (23.4%)12(40.0%)34Boerboel7 (7.4%)4 (13.3%)11Bullmastiff11 (11.7%)0 (0.0%)11German Shepherd13 (13.8%)2 (6.7%)15Caucasian Shepherd8 (8.5%)0 (0.0%)8Crossbreed15(16.0%)9 (30.0%)24Poodle5 (5.3%)1 (3.3%)6Rottweiler7 (7.4%)1 (3.3%)8Maltese2 (2.1%)0 (0.0%)2Shih Tzu1 (1.1%)0 (0.0%)1Alsatian Shepherd1 (1.1%)0 (0.0%)1American Bully1 (1.1%)0 (0.0%)1Belgium Malinois1 (1.1%)1 (3.3%)2Total94 (100.0%)30 (100.0)124 (100%)Only CPV-positive dogs that received treatment were included. Dogs that did not receive treatment or died before treatment were excluded Table 2Descriptive distribution of CPV-positive dogs by service type and breed (n = 124)BreedInpatient N (%)Outpatient N (%)TotalLocal Breed22 (23.4%)12(40.0%)34Boerboel7 (7.4%)4 (13.3%)11Bullmastiff11 (11.7%)0 (0.0%)11German Shepherd13 (13.8%)2 (6.7%)15Caucasian Shepherd8 (8.5%)0 (0.0%)8Crossbreed15(16.0%)9 (30.0%)24Poodle5 (5.3%)1 (3.3%)6Rottweiler7 (7.4%)1 (3.3%)8Maltese2 (2.1%)0 (0.0%)2Shih Tzu1 (1.1%)0 (0.0%)1Alsatian Shepherd1 (1.1%)0 (0.0%)1American Bully1 (1.1%)0 (0.0%)1Belgium Malinois1 (1.1%)1 (3.3%)2Total94 (100.0%)30 (100.0)124 (100%)Only CPV-positive dogs that received treatment were included. Dogs that did not receive treatment or died before treatment were excluded Descriptive distribution of CPV-positive dogs by service type and breed (n = 124) Only CPV-positive dogs that received treatment were included. Dogs that did not receive treatment or died before treatment were excluded Overall treatment outcome by CPV status Across the entire cohort, most dogs received treatment and recovered (n = 166, 78.67%). Among the 129 CPV-positive dogs, two died before treatment could be administered at the facility, three did not receive treatment at the facility, and one was lost to follow-up, resulting in incomplete treatment outcome data for six cases. Among the remaining CPV-positive dogs, 70.54% (91/129) recovered following treatment, whereas 24.81% (32/129) died despite treatment. In contrast, recovery was higher among CPV-negative dogs, with 91.46% (75/82) recovering and 8.54% (7/82) dying following treatment (Table 3). Table 3Distribution of canine parvovirus diagnoses by treatment status and clinical outcome (n = 211)Treatment OutcomeCPV+CPV-TotalTreated and Recovered91(70.54%)75(91.46%)166(78.67%)Treated and Deceased32(24.81%)7(8.54%)39(18.48%)Incomplete Treatment Regimen1(0.78%)0(0.00%)1(0.47%)Deceased Before Treatment2(1.55%)0(0.00%)2(0.95%)No Treatment3(2.33%)0(0.00%)3(1.42%)Total129(100.00%)82(100.00%)211(100%) Table 3Distribution of canine parvovirus diagnoses by treatment status and clinical outcome (n = 211)Treatment OutcomeCPV+CPV-TotalTreated and Recovered91(70.54%)75(91.46%)166(78.67%)Treated and Deceased32(24.81%)7(8.54%)39(18.48%)Incomplete Treatment Regimen1(0.78%)0(0.00%)1(0.47%)Deceased Before Treatment2(1.55%)0(0.00%)2(0.95%)No Treatment3(2.33%)0(0.00%)3(1.42%)Total129(100.00%)82(100.00%)211(100%) Distribution of canine parvovirus diagnoses by treatment status and clinical outcome (n = 211) 91 (70.54%) 75 (91.46%) 166 (78.67%) 32 (24.81%) 7 (8.54%) 39 (18.48%) 1 (0.78%) 0 (0.00%) 1 (0.47%) 2 (1.55%) 0 (0.00%) 2 (0.95%) 3 (2.33%) 0 (0.00%) 3 (1.42%) 129 (100.00%) 82 (100.00%) 211 (100%) Clinical signs and measurements associated with canine parvovirus infection Univariable logistic regression analysis revealed that appetite status (p < 0.001), vomiting (p < 0.001), faecal consistency (p < 0.001), the smell of faeces (p < 0.001), diarrhoea status (p < 0.001), and body weight (0.030) were significantly associated with increased odds of CPV diagnosis. The body temperature of CPV-positive dogs ranged from 36.7 °C to 41.2 °C (mean ± SD = 38.9 ± 0.78 °C), while their body weight ranged from 0.8 kg to 37.0 kg (median [IQR] = 6.2[4.4–9.9]). Lethargy and body temperature were not significantly associated with CPV diagnosis in this population (Table 4). Table 4Clinical signs and measurements associated with CPV diagnosis in dogs that presented to the facility (n = 211)VariableCPV+CPV-Totalp-valueOdds ratio95%CI Lethargy Noa171633 Yes112661780.2201.600.76–3.37 Appetite status Normala4913< 0.001Ref Complete loss of appetite86261120.0027.442.12–26.15 Reduced appetite3947860.3301.870.53–6.53 Vomiting status Noa253762 Yes10445149< 0.0013.421.85–6.34 Consistency of faeces Solida153045< 0.001Ref Watery9431125< 0.0016.072.89–12.72 Semi-solid2021410.1501.910.80–4.55 Smell of faeces Normala315182 Foul smell9831129< 0.0015.202.85–9.50 Diarrhoea status No diarrhoea304878< 0.001Ref Non-Haemorrhagica4224660.0032.801.42–5.52 Haemorrhagic571067< 0.0019.124.05–20.55 Weight in kg N129820.0300.960.92-1.00Minimum0.81.1Maximum37.035.5Median (IQR)6.2(4.4–9.9)10.5(5.1–14.2) Body Temperature (°C) N129820.6500.920.65–1.30Minimum36.736.9Maximum41.241.5Mean ± SD38.9 ± 0.7839.0 ± 0.85CI = confidence interval; Ref/a=reference variable; SD=Standard deviation; IQR=Interquartile range Table 4Clinical signs and measurements associated with CPV diagnosis in dogs that presented to the facility (n = 211)VariableCPV+CPV-Totalp-valueOdds ratio95%CI Lethargy Noa171633 Yes112661780.2201.600.76–3.37 Appetite status Normala4913< 0.001Ref Complete loss of appetite86261120.0027.442.12–26.15 Reduced appetite3947860.3301.870.53–6.53 Vomiting status Noa253762 Yes10445149< 0.0013.421.85–6.34 Consistency of faeces Solida153045< 0.001Ref Watery9431125< 0.0016.072.89–12.72 Semi-solid2021410.1501.910.80–4.55 Smell of faeces Normala315182 Foul smell9831129< 0.0015.202.85–9.50 Diarrhoea status No diarrhoea304878< 0.001Ref Non-Haemorrhagica4224660.0032.801.42–5.52 Haemorrhagic571067< 0.0019.124.05–20.55 Weight in kg N129820.0300.960.92-1.00Minimum0.81.1Maximum37.035.5Median (IQR)6.2(4.4–9.9)10.5(5.1–14.2) Body Temperature (°C) N129820.6500.920.65–1.30Minimum36.736.9Maximum41.241.5Mean ± SD38.9 ± 0.7839.0 ± 0.85CI = confidence interval; Ref/a=reference variable; SD=Standard deviation; IQR=Interquartile range Clinical signs and measurements associated with CPV diagnosis in dogs that presented to the facility (n = 211) CI = confidence interval; Ref/a=reference variable; SD=Standard deviation; IQR=Interquartile range In the multivariable logistic regression model, appetite status, vomiting, diarrhoea status, and body weight were independently associated with CPV diagnosis. Dogs with complete loss of appetite had significantly increased odds of CPV diagnosis compared with dogs with normal appetite (OR = 4.29, 95% CI: 1.11–16.63; p = 0.035), whereas reduced appetite was not statistically significant. Vomiting was associated with 2.75-fold higher odds of CPV diagnosis (OR = 2.75, 95% CI: 1.35–5.59; p = 0.005). Each kilogram increase in body weight was associated with a 6% reduction in the odds of CPV diagnosis (OR = 0.94, 95% CI: 0.90–0.99; p = 0.010). Regarding diarrhoea status, dogs with haemorrhagic diarrhoea had significantly higher odds of CPV diagnosis compared with dogs without diarrhoea (OR = 6.84, 95% CI: 2.76–16.98; p < 0.001), while non-haemorrhagic diarrhoea did not remain statistically significant in the adjusted model. The final model demonstrated acceptable calibration (Hosmer–Lemeshow χ² = 11.64, p = 0.168), was statistically significant overall (Omnibus test χ² = 62.55, p < 0.001), and had moderate explanatory power (Nagelkerke R² = 0.348). Model discrimination was good, with an area under the receiver operating characteristic curve (AUC) of 0.80 (95% CI: 0.73–0.86) (Table 5). Table 5Multivariable logistic regression analysis of clinical signs and measurements associated with canine parvovirus diagnosis in dogs (n = 211)Variables in the Equationp-valueOdds Ratio95% CIAppetite Status0.003REFNormalaReduced Appetite0.5691.480.39–5.68Complete Loss of Appetite0.0354.291.11–16.63Vomitinga0.0052.751.35–5.59Diarrhoea Status< 0.001REFNo Diarrhoea aNon-haemorrhagic diarrhoea0.0751.980.93–4.19Haemorrhagic Diarrhoea< 0.0016.842.76–16.98Weight (kg)0.0100.940.90–0.99Model fit: Nagelkerke R2 = 0.348; Hosmer–Lemeshow p value = 0.168ᵃ Reference categories: normal appetite (for appetite status), no diarrhoea (for diarrhoea status), and. no vomiting (for vomiting). Reference categories were retained from the univariable regression for the multivariable regression analysis Table 5Multivariable logistic regression analysis of clinical signs and measurements associated with canine parvovirus diagnosis in dogs (n = 211)Variables in the Equationp-valueOdds Ratio95% CIAppetite Status0.003REFNormalaReduced Appetite0.5691.480.39–5.68Complete Loss of Appetite0.0354.291.11–16.63Vomitinga0.0052.751.35–5.59Diarrhoea Status< 0.001REFNo Diarrhoea aNon-haemorrhagic diarrhoea0.0751.980.93–4.19Haemorrhagic Diarrhoea< 0.0016.842.76–16.98Weight (kg)0.0100.940.90–0.99Model fit: Nagelkerke R2 = 0.348; Hosmer–Lemeshow p value = 0.168ᵃ Reference categories: normal appetite (for appetite status), no diarrhoea (for diarrhoea status), and. no vomiting (for vomiting). Reference categories were retained from the univariable regression for the multivariable regression analysis Multivariable logistic regression analysis of clinical signs and measurements associated with canine parvovirus diagnosis in dogs (n = 211) Model fit: Nagelkerke R2 = 0.348; Hosmer–Lemeshow p value = 0.168 ᵃ Reference categories: normal appetite (for appetite status), no diarrhoea (for diarrhoea status), and. no vomiting (for vomiting). Reference categories were retained from the univariable regression for the multivariable regression analysis Risk analysis on the basis of CPV diagnosis and service rendered Compared with CPV-negative dogs, CPV-positive dogs were 7.57 times more likely to be managed as inpatients (p < 0.001), and this association was statistically significant. (Table 6). The single outpatient (local dog) with an incomplete treatment regimen was retained in this analysis as the service was rendered; however, dogs that received no treatment or were deceased before treatment were excluded from the comparison. Table 6Univariate logistic regression analysis of services rendered in relation to canine parvovirus diagnosis (n = 206)VariablePositiveNegativeTotalPrevalence%p-valueOdds ratioLower 95%CIUpper 95%CI Service Rendered (Inpatient/Outpatient) Outpatienta30588834.1%< 0.0017.574.0414.20Inpatient942411879.7%Total12482206CI = confidence interval; a = reference variable; excluding dogs that received no treatment and death before treatment Table 6Univariate logistic regression analysis of services rendered in relation to canine parvovirus diagnosis (n = 206)VariablePositiveNegativeTotalPrevalence%p-valueOdds ratioLower 95%CIUpper 95%CI Service Rendered (Inpatient/Outpatient) Outpatienta30588834.1%< 0.0017.574.0414.20Inpatient942411879.7%Total12482206CI = confidence interval; a = reference variable; excluding dogs that received no treatment and death before treatment Univariate logistic regression analysis of services rendered in relation to canine parvovirus diagnosis (n = 206) CI = confidence interval; a = reference variable; excluding dogs that received no treatment and death before treatment Association between type of service rendered to positive CPV cases by sex, age group, and breed Inpatient management was more common than outpatient care among CPV-positive dogs. In univariate analysis, female dogs (p = 0.069), dogs aged ≤ 12 weeks (p = 0.144) and 13–24 weeks (p = 0.170), and foreign breeds (p = 0.080) showed higher odds of inpatient care compared to their respective reference groups; however, these associations were not statistically significant (Table 7). Table 7Association between sex, age group, breed (general) and type of service rendered in canine parvovirus-positive cases (n = 124; excluding dogs that received no treatment and death before treatment)Service RenderedPredisposing factorInpatient (%)Outpatient (%)Totalp-valueOdds RatioLower CIUpper CI Sex Male51 (54.3%)22 (73.3%)73RefFemale43 (45.7%)8 (26.7%)510.0692.320.945.73 Age Group More than 24 weeks11 (11.7%)7 (23.3%)180.300RefLess than or equal to 12 weeks41 (43.6%)11 (36.7%)520.1442.370.757.5513 to 24 weeks42 (44.7%)12 (40.0%)540.1702.230.717.00 Breed Local Breed22 (23.4%)12 (40.0%)34Foreign Breed72 (76.6%)18 (60.0%)900.0802.180.915.22CI = confidence interval; Ref/ a = reference variable; excluding dogs that received no treatment and death before treatment Table 7Association between sex, age group, breed (general) and type of service rendered in canine parvovirus-positive cases (n = 124; excluding dogs that received no treatment and death before treatment)Service RenderedPredisposing factorInpatient (%)Outpatient (%)Totalp-valueOdds RatioLower CIUpper CI Sex Male51 (54.3%)22 (73.3%)73RefFemale43 (45.7%)8 (26.7%)510.0692.320.945.73 Age Group More than 24 weeks11 (11.7%)7 (23.3%)180.300RefLess than or equal to 12 weeks41 (43.6%)11 (36.7%)520.1442.370.757.5513 to 24 weeks42 (44.7%)12 (40.0%)540.1702.230.717.00 Breed Local Breed22 (23.4%)12 (40.0%)34Foreign Breed72 (76.6%)18 (60.0%)900.0802.180.915.22CI = confidence interval; Ref/ a = reference variable; excluding dogs that received no treatment and death before treatment Association between sex, age group, breed (general) and type of service rendered in canine parvovirus-positive cases (n = 124; excluding dogs that received no treatment and death before treatment) CI = confidence interval; Ref/ a = reference variable; excluding dogs that received no treatment and death before treatment Discussion Canine parvovirus remains a significant clinical challenge in small animal practice worldwide [1, 17, 21]. This study identifies clinical characteristics observed at presentation that are associated with a positive CPV diagnosis, service-rendered dynamics, and outcomes in affected dogs. Classic signs such as anorexia, vomiting, and haemorrhagic diarrhoea were strongly associated with CPV positivity, which is consistent with the established enteric syndrome of the virus [12, 18, 23, 27]. Haemorrhagic diarrhoea showed the highest odds of association with canine parvovirus diagnosis (OR = 6.84, 95% CI: 2.76–16.98; p < 0.001) among the predictors retained in the multivariable model, followed by complete loss of appetite (OR = 4.29, 95% CI: 1.11–16.63; p = 0.035) and vomiting (OR = 2.75, 95% CI: 1.35–5.59; p = 0.005). These findings potentially highlight that these clinical signs may support routine case assessment, particularly among unvaccinated or high-risk dogs. Early recognition of such signs can facilitate supportive management, enhance prognostic evaluation, and guide timely isolation measures to limit disease transmission within hospital and community settings. Conversely, not all confirmed cases presented with these hallmark clinical signs at the time of admission; in some instances, patients lacked haemorrhagic diarrhoea, showed no vomiting, were not anorexic, and had no diarrhoea. These findings, consistent with reports in other areas, highlight that the absence of individual signs does not exclude CPV, underscoring the need for a thorough assessment of suspected cases [7, 18, 27]. Similar patterns were observed in Ghana by Folitse et al. [12] and in Egypt by Sayed-Ahmed et al. [37], where a significant proportion of dogs with non-haemorrhagic diarrhoea tested positive. Clinicians should therefore consider CPV as a strong differential in young dogs presenting with diarrhoea, as the haemorrhagic nature typically emerges with disease progression and severity, while earlier stages may involve non-haemorrhagic diarrhoea of variable colour and consistency [5, 30]. Body weight also emerged as a significant independent predictor (OR = 0.94, 95% CI: 0.90–0.99; p = 0.010), reflecting the heightened susceptibility of puppies and young dogs, which generally weigh less than adults do [26]. Although the final multivariable model excluded foul-smelling diarrhoea (OR = 5.20, 95% CI: 2.85–9.50; p < 0.001), this sign remains clinically relevant, having been observed in a substantial number of cases and reflecting the severe mucosal damage characteristic of CPV enteritis [28]. Rectal temperature measured at initial case presentation varied widely, ranging from marked hyperthermia (41.2 °C) to hypothermia (36.7 °C), with a mean presentation temperature of 38.9 °C. While fever is often associated with acute CPV, temperature was not statistically significant in this study, likely reflecting the variability of clinical presentation of the disease. Early hyperthermia may transition to hypothermia as vomiting and diarrhoea progress, driven by dehydration and circulatory compromise [30]. Similar to our findings, Godsall et al. [14] and Miranda et al. [26] reported that most CPV cases were normothermic at presentation, corroborating the limited predictive value of body temperature. Collectively, these findings emphasise the heterogeneity of CPV at clinical presentation, which may vary with disease stage, viral strain, vaccination status, age, and breed predisposition [3, 9, 21, 25]. Nevertheless, the observed associations suggest that these clinical signs may contribute to the clinical assessment of CPV. Even in cases with variable presentation patterns, these symptoms may remain clinically useful for guiding presumptive diagnosis while awaiting confirmatory antigen testing. Further reflecting the severity of CPV presentations, our analysis revealed that CPV-positive dogs were almost eight times more likely to require hospitalisation than were CPV-negative dogs (OR = 7.57, 95% CI: 4.04–14.20; p = < 0.001), highlighting the condition’s intensity and greater need for inpatient care [7, 18, 34]. The breed distribution among the inpatient and outpatient groups varied; local breeds and crossbreeds accounted for more than 60% of the outpatient cases, whereas other pure breeds, such as Rottweilers, Bullmastiffs, German Shepherds, and Caucasian Shepherds, were mostly treated as inpatients. This pattern may partly reflect socio-economic influences on case management decisions, as owners of pure or foreign breeds may be more willing or financially able to pursue prolonged inpatient care compared with owners of local or crossbreed dogs. Similar observations have been reported in veterinary practice settings where breed ownership may correlate with differences in economic status and healthcare-seeking behaviour [2, 4, 20]. Additionally, females and dogs younger than 24 weeks of age received the most inpatient care. The pattern of dogs under 24 weeks of age (puppies < 6 months) requiring inpatient care may be due to the severe nature of CPV in this particular age group, due to their immature immunity and increased susceptibility to systemic complications [18, 32, 35]. Collectively, breed and age-related factors may influence hospitalisation decisions for dogs with CPV, as some pure breeds or younger age groups may present with more severe symptoms and disease progression [3, 27, 30, 33]. The overall mortality rate among CPV-positive dogs in this study was 26.4%, including the two dogs that died before treatment, underscoring the substantial risk CPV poses to the canine population in clinical settings. Without treatment, mortality can reach as high as 91% in puppies, highlighting the lifesaving importance of early recognition and timely intervention [28]. The mortality rate reported in this study is lower than that reported in Nigeria (72% by Ukwueze et al. [40] and 80.6% by Ogbu et al. [32] and Cameroon, where Mebanga et al. [24] reported 58.9%, but it is comparable to that reported by Godsall et al. [14] in the United Kingdom (30.2%). In contrast, it is higher than that reported by Chalifoux et al. [7] in the United States (9%) and Miranda et al. (2016) in Portugal (16.3%). These differences may be due to variations in disease severity at presentation, tertiary-level care, diagnostic and therapeutic capacities, demographic characteristics, vaccination, and the prevalence of regional comorbid conditions such as helminth infections and ectoparasite infestations [3, 11, 34, 35, 42]. Notably, all three dogs that received no treatment were CPV positive, and one was subsequently lost to follow-up, leading to an incomplete treatment regimen, likely reflecting the financial or logistical constraints this disease imposes on dog owners [19, 20]. This finding may reflect how such barriers can affect outcomes and points to the broader socioeconomic consequences of CPV [36]. For CPV-negative patients, which comprised dogs presenting with similar clinical signs of acute gastroenteritis, the mortality rate was relatively low, at approximately 8.5%, demonstrating the considerably greater risk associated with CPV than with the other conditions examined, which is consistent with other studies [14, 33]. Limitations Limitations should be considered when interpreting the findings of this study. Firstly, the aggregate CPV-negative cases, which presented with clinical signs similar to those of CPV-positive cases, could have been due to several possible causes of acute gastroenteritis, limiting the comparison. In addition, immunochromatographic antigen assays may occasionally result in false-negatives compared to molecular-based diagnostic methods due to reduced antigen shedding at certain stages of CPV infection [10]. Laboratory indicators of disease severity, such as hydration status and anaemia, including complete blood count, serum biochemistry, and electrolyte profiling, were not consistently available or standardised across all enrolled dogs. Additionally, objective clinical severity scoring was not uniformly applied at presentation, and treatment protocols may have varied among attending clinicians, potentially influencing clinical outcomes. Concurrent infections with other pathogens or underlying health conditions were not systematically investigated and may have influenced clinical presentation, disease severity, and outcomes observed among some cases. Additionally, recall bias could have affected the accuracy of some clinical data, as information on vaccination and clinical signs at presentation partly relied on owner recall. Although multivariable analyses adjusted for various confounders, residual bias may still exist, including clinician-driven treatment decisions or financial limitations of dog owners. The sample size, while adequate for the primary analysis, was constrained by the number of eligible cases within the study period, limiting generalisability. Inpatient management at the study facility was based on 12-hour hospitalisation without continuous 24-hour care, which may partly explain outcome differences compared to hospitals in high-income countries. Survival time analyses and related factors were beyond the scope of this current study and will be addressed in subsequent work. Lastly, the allocation of inpatient versus outpatient care was not performed within a randomised controlled setting, complicating direct causal comparisons between service types. Conclusion The findings of this study reaffirm that canine parvovirus is a preventable yet persistent threat to unvaccinated or incompletely vaccinated puppies. Classic signs, including haemorrhagic diarrhoea, vomiting, anorexia and lower body weight, were independently predictive of positive CPV cases. Among dogs presenting with comparable clinical signs, CPV-positive cases were more likely to require hospitalisation and experienced higher mortality than CPV-negative cases, reflecting the severity of the disease. Strengthening vaccination coverage, client education, and access to supportive care are crucial to reducing CPV-related deaths and improving canine health outcomes in Ghana.