Abstract
OBJECTIVE
To evaluate factors contributing to owner decision-making, satisfaction, and perception of quality of life (QOL) with treatment of dogs with nontraumatic hemoabdomen (NTH).
ANIMALS
132 client-owned dogs.
PROCEDURES
An electronic survey was administered to owners of 436 dogs that presented emergently with NTH to a single institution between January 2015 and May 2022. Following survey response, retrospective data collection was performed.
RESULTS
Owners reported QOL as the most important factor influencing their decision-making (92%), followed by risk of cancer (57%) or time remaining with their pet (56%). QOL scores were significantly higher with surgery versus those with palliative care (P = .007). Median survival time (MST) was 213 days with surgery and 39 days with palliative care (P = .049). Survival benefit of surgery was lost when considering only dogs with malignant histopathology (MST, 81 days; P = .305). Owners were more likely to be satisfied when they chose surgery over either euthanasia or palliative care (P = .039). Thirty-four owners (26%) second-guessed or were unsure of their decision.
CLINICAL RELEVANCE
Surgery resulted in the longest MST with greater perceived QOL and owner satisfaction compared with both palliative care and euthanasia and should therefore be considered highly. The importance of malignancy and survival time on owners’ decisions, along with the negative impact of metastasis on survival, underscores the importance of timely preoperative staging. The rate of second-guessing highlights the need for standardization of NTH discussions including treatment options and potential outcomes in order to effectively and efficiently guide treatment of patients with this common presentation.
Introduction
Nontraumatic hemoabdomen or hemoperitoneum (NTH) is a common, life-threatening canine emergency. Dog owners are faced with the emotionally and financially demanding decision of whether to pursue urgent medical or surgical intervention despite an unknown underlying cause and often guarded prognosis.1 Common causes of NTH include benign or malignant intra-abdominal neoplasia, hematoma, organ torsion, gastric dilatation-volvulus, and acquired or congenital coagulopathies.1–3 NTH due to organ torsion or coagulopathy is more quickly elucidated and can have a good long-term prognosis with proper treatment; however, the majority of cases carry an uncertain diagnosis and prognosis at initial workup.1–3
In the remaining cases, a bleeding abdominal mass is often suspected. Hemorrhage can originate from hepatic, renal, adrenal, prostatic, and retroperitoneal masses, but most frequently occurs from splenic masses.3–5 Multiple studies have shown malignant neoplasia is the most common cause of NTH in dogs, occurring at a rate of 68% to 87% compared with benign or nonneoplastic etiologies.2 Splenic hemangiosarcoma is identified in 63% to 70% of NTH cases and carries a poor long-term prognosis, with < 10% of dogs surviving 1 year.2,3,5
A variety of point-of-care tests and advanced diagnostics have been studied as negative prognostic indicators for NTH.5–11 Confidently reaching a diagnosis prior to tissue acquisition is not currently feasible, and definitive diagnosis relies on histopathologic evaluation.1,2,6,12 Considering an uncertain prognosis at presentation and the prevalence of malignancy in NTH cases, ethically and clinically sound treatment remains controversial and is ultimately guided by discussions between veterinarians and clients, which may vary greatly among individuals depending on clinical experience and personal biases.1–3,5,6,13
A previous study12 of dogs with suspected neoplastic hemoperitoneum that underwent surgery explored factors influencing dog owners’ decision to pursue surgery, surgical complications, and patient quality of life (QOL) following discharge. However, a limitation of this study was selection for neoplastic hemoabdomen only, as the presence of neoplasia is typically unconfirmed when the decision to pursue surgery is made.1,12 A recent study13 illustrated that the rate of euthanasia of NTH cases is significantly impacted by veterinarian experience level, highlighting the influential role of veterinarian-client communication in this pivotal decision. To the authors’ knowledge, there have been no large retrospective studies evaluating owner perception and satisfaction of outcome in patients with NTH of benign and malignant causes whose owners elected immediate euthanasia, palliative care, or surgery.
The primary objective of this study was to evaluate the factors that contribute to owner decision-making at the time of emergent presentation, owner satisfaction in relation to the decision made, and owner perception of the pet’s QOL before, during, and after treatment of dogs with NTH. Through investigation of these factors, evidence-based recommendations and expectations can be provided to inform and facilitate discussions with clients on treatment options for dogs with NTH.
Materials and Methods
Study design and initial case identification criteria
A retrospective cohort study design was utilized for identification and evaluation of patients with nontraumatic, spontaneous hemoabdomen (NTH). Medical records of dogs that presented with NTH between January 2015 and May 2022 at the Cornell University Hospital for Animals, predominantly through the Emergency Service, were preliminarily reviewed. NTH was defined by presence of hemorrhagic peritoneal effusion, confirmed by abdominocentesis and clinicopathologic fluid analysis. Cases were excluded if hemoperitoneum was secondary to trauma, iatrogenic causes (eg, recent surgery), coagulopathy, gastric dilatation-volvulus, liver lobe torsion, or splenic torsion due to a more rapidly elucidated etiology and predictable prognosis without the need for histopathology. Patients were included irrespective of case outcome (eg, immediate euthanasia, palliative care, surgery). Client and patient name as well as client email addresses were recorded for dogs meeting the initial inclusion criteria. Cases without a client email address on file were excluded.
Survey
An owner survey was created based on the study objectives to obtain appropriate patient follow-up and outcome. The survey was adapted from previous veterinary investigations of QOL and included multiple-choice, yes-or-no, and open-ended questions.12,14,15 The survey was hosted on Qualtrics, with inherent logic present allowing for a unique user experience. The study design and survey were submitted to the Institutional Review Board for Human Participants, and the protocol was granted exemption (protocol #IRB0145507). The complete survey (Supplementary Appendix S1) was distributed via email on July 29, 2022, and was open for 2 months following initial dissemination. Survey response was incentivized by entry into a raffle drawing. Dogs whose owners completed the survey met inclusion criteria for further clinical record review. Dogs whose owners either partially completed the survey or did not respond were excluded.
Clinical records review
A complete medical record review was performed for all dogs whose owners responded to the survey. Additional cases were excluded if upon deeper review previous inclusion criteria were not met. Treatment decision outcome was classified into the following categories: (1) no surgical intervention—natural death or humane euthanasia during initial visit or within 24 hours of discharge; (2) no surgical intervention—alive at discharge with palliative care elected; (3) surgical intervention—euthanized or died during procedure or during initial hospitalization; and (4) surgical intervention—alive at discharge.
Data obtained from medical records included signalment, body weight (kg), body condition score, heart rate (tachycardia > 140 beats/min), mean arterial pressure, and relevant lab work (PCV, total solids, platelet count, hematocrit, lactate) from initial presentation. Additional data collected consisted of admitting clinician experience (intern, resident, or faculty), thoracic focused assessment with sonography for trauma (normal or abnormal [free fluid, B-lines, absent glide sign]), thoracic radiographs (yes [normal or metastatic lung pattern] or no), blood transfusion (yes or no), advanced abdominal imaging (ultrasound, CT, or both), surgical information (time to surgery and surgical outcome), days hospitalized, and cost and outcome of initial hospital visit.
Further data obtained included reported sources of hemorrhage based on imaging and surgery, metastasis based on imaging and surgery, histopathology results if available, medications or supplements given following discharge (including aminocaproic acid, Yunnan Baiyao, Coriolus versicolor polysaccharopeptide [I’m-Yunity], and chemotherapy), and overall survival time. Overall survival time was defined in days from initial presentation to death. Death was verified through the medical record, owner survey response, or both. If diagnosis of hemangiosarcoma was confirmed with histopathology, mitotic score and clinical stage were noted. Clinical stage was assigned based on the modified World Health Organization scheme.16 Stage I was hemangiosarcoma confined to spleen, stage II was ruptured splenic hemangiosarcoma, and stage III was clinically detectable distant metastases or concurrent right atrial mass.16
Statistical analysis
Continuous data were assessed for normality with the Shapiro-Wilk test. Categorical data were reported as frequency (%). Normally distributed continuous data were reported as mean ± SE, and nonnormally distributed continuous data were reported as median and IQR. Categorical comparisons were performed utilizing a χ2 test or a Fisher exact test when cell value was ≤ 5. Comparisons between nonparametric continuous data and dichotomous variables were performed using the rank sum test. Comparisons between nonparametric continuous data and polychotomous variables were performed using the Kruskal-Wallis test. Odds ratios and 95% CI are reported where warranted. The equality of medians with regard to owner-reported influences on decision-making was tested using a Wilcoxon signed rank test. QOL scores were compared between palliative and surgery utilizing the rank sum test.
Univariable Cox regression estimates were performed to evaluate survival time from treatment decision for all factors. Kaplan-Meier survival curves were generated for all patients and for patients undergoing surgery versus palliative care. Hazard ratios with 95% CIs and P values were reported. Patients that were alive at the end of study period or known to have died from a cause unrelated to the hemoabdomen were right censored. Values of P < 0.05 were considered significant. All statistical calculations were performed with commercial software (Stata version 15.1; StataCorp LLC).
Results
Demographics
Medical record query revealed 710 records with a diagnosis of hemoabdomen at the Cornell University Hospital for Animals January 2015 to May 2022. Of these dogs, 503 were identified to have NTH on preliminary review. Email addresses were available for 436 dogs. The survey was distributed to this group, with 22 undeliverable emails. This resulted in distribution of 414 unique surveys. Surveys were started by 165 owners and 156 were fully completed, for a completion rate of 94.5% and overall response rate of 37.7%. Median time to complete the survey was 411 seconds (range, 312 to 758 seconds). Complete record review was performed after survey responses, resulting in exclusion of 24 cases not meeting inclusion criteria. A total of 132 unique responses were evaluated, for an overall survey yield of 31.9%.
The most common purebred dogs represented were Golden Retriever (13/132 [9.9%]), Labrador Retriever (12/132 [9.1%]), and German Shepherd Dog (10/132 [7.6%]). Most dogs were mixed breed (45/132 [34.1%]). Median age at presentation was 10.2 years (n = 132; range, 9.0 to 11.6 years). Mean weight was 28.8 ± 1.05 kg (n = 125). Of the 132 dogs, 57 (43.2%) were male and 75 (56.8%) were female. One (1.3%) female was intact and 4 (7.0%) males were intact, while the remaining 127 (96%) dogs were spayed or neutered. Average body condition score was 5.5 ± 0.12 out of 9 (n = 77).
Preoperative evaluation
On presentation, the patient was evaluated primarily by an intern in 64 of 132 (48.5%) cases or by a resident or faculty member in 68 of 132 (51.5%) cases. Mean presenting heart rate was 144.2 ± 2.6 beats/min (n = 128). Dogs were classified as tachycardic in 74 of 128 (57.8%) cases. Mean PCV was 32.3 ± 0.8% (n = 119). Mean total solids were 6.3 ± 0.09 g/dL (n = 118). Median platelet count was 134,200/µL (n = 43; range, 61,000 to 189,000). Average mean arterial pressure was 95.8 ± 4.0 mm Hg (n = 54). Mean serum lactate concentration was 5.0 mmol/L (n = 104; range, 2.5 to 7.6 mmol/L). A total of 41 of 132 (31.1%) patients underwent a packed RBC transfusion. Out of the 41 patients that received a packed RBC blood transfusion, 26 (63.4%) underwent surgery, 5 (12.2%) received palliative care, and 10 (24.4%) were euthanized. A total of 26 of 40 (65%) dogs that underwent surgery received a blood transfusion, 5 of 22 (22.7%) dogs that underwent palliative care received a blood transfusion, and 10 of 70 (14.3%) dogs that were euthanized within 24 hours received a blood transfusion.
Thoracic focused assessment with sonography for trauma was performed in 69 of 132 (52.3%) dogs, with abnormalities identified in 14 of 69 (20.3%) patients. Abnormalities included pleural effusion in 6 of 14 (46.2%) patients and B-lines in 8 of 14 (61.5%) patients. One patient was noted to have possible pericardial effusion. Thoracic radiographs were performed in 63 of 132 (47.7%) patients. Of these 63 patients, 44 (69.8%) had no significant findings, 7 (11.1%) had evidence of metastasis, and 7 (11.1%) had other pulmonary findings not suspected to be metastasis.
Abdominal imaging was performed in 64 of 132 (48.5%) dogs, including abdominal ultrasound in 57 of 64 (89.1%) dogs and CT in 5 of 64 (7.8%) dogs, with 2 dogs undergoing both ultrasound and CT. The source of hemorrhage based on imaging was suspected to be spleen in 42 of 64 (65.6%) cases, liver in 19 of 64 (29.7%) cases, retroperitoneal in 1 of 64 (1.6%) cases, adrenal in 1 of 64 (0.8%) cases, and unknown/undetermined in 5 of 64 (3.8%) cases. Metastasis was strongly suspected on abdominal imaging in 23 of 64 (35.9%) dogs, potentially suspected in 20 of 64 (31.3%) dogs, and not identified in 21 of 64 (32.8%) dogs.
When presumptive metastasis was identified during staging, 18 of 44 (40.9%) patients underwent surgery, 12 of 44 (27.3%) patients underwent palliative care, and 14 of 44 (31.8%) patients were euthanized within 24 hours. When confident metastasis was identified during staging, 6 of 25 (21.7%) patients underwent surgery, 6 of 25 (26.1%) patients received palliative care, and 13 of 25 (52.2%) patients were euthanized within 24 hours.
Outcomes
A total of 70 of 132 (53.0%) dogs either died or were euthanized within 24 hours of presentation, 22 of 132 (16.7%) dogs were discharged for palliative care and survived > 24 hours, and 40 of 132 (30.3%) dogs underwent surgery. Of 40 dogs that underwent surgery, 4 (10%) either were euthanized or died while in the hospital perioperatively. Two dogs were euthanized intraoperatively, 1 dog was euthanized postoperatively, and 1 dog died postoperatively. Median time from hospital admission until surgery was 16 hours (n = 39; range, 8 to 24 hours). Median time of hospitalization for patients undergoing surgery was 3 days (n = 40; range, 2 to 3 days), which was longer than the median hospitalization time for patients receiving palliative care (n = 22; range, 0 to 1 days; P < .001).
Intraoperatively, the primary source of hemorrhage was identified as the spleen in 37 of 40 (92.5%) cases, liver in 2 of 40 (5%) cases, and kidney in 1 of 40 (2.5%) cases. When compared with surgery, location of hemorrhage on preoperative imaging was classified correctly in 97.22% of cases. Intraoperative metastasis was confidently identified in 11 of 40 (27.5%) patients, presumptive based on appearance in 4 of 40 (10%) patients, and not identified in 25 of 40 (62.5%) patients.
Histopathology was available for 41 cases (40 surgical biopsies, 1 postmortem biopsy), of which 31 (75.6%) were considered malignant disease. Malignant diseases were categorized as hemangiosarcoma in 26 of 41 (83.9%) cases, splenic leiomyosarcoma in 2 of 41 (4.9%) cases, hepatocellular carcinoma in 1 of 41 (2.4%) cases, splenic fibrohistiocytic nodule in 1 of 41 (2.4%) cases, and splenic liposarcoma in 1 of 41 (2.4%) cases. Hemangiosarcoma was localized to spleen in 24 of 26 (92%) cases, liver in 1 of 26 (4%) cases, and kidney in 1 of 26 (4%) cases. Median mitotic count in 10 hpf was 11 (n = 29; range, 5 to 21). For patients with hemangiosarcoma, 10 of 24 (41.7%) patients were assigned a mitotic score of 0, 6 of 24 (25.0%) a score of 1, 4 of 24 (16.7%) a score of 2, and 4 of 24 (16.7%) a score of 3. Clinical stage was recorded in all 26 dogs, with 16 of 26 (61.5%) dogs having stage 2 and 10 of 26 (38.5%) dogs having stage 3 hemangiosarcoma. Patients were classified as having benign disease in 10 of 41 (24.4%) cases, with all cases diagnosed with splenic hematoma as the NTH cause. Seven of 10 (70%) cases had concurrent nodular hyperplasia.
Dogs were administered additional hemostatic or antifibrinolytic medications or supplements either during hospitalization or after discharge. A total of 58 of 132 (43.9%) dogs were administered aminocaproic acid while hospitalized. Of the 58 patients that survived to discharge, 26 (44.8%) were discharged with additional medications or supplements. Fifty-six owners responded to the survey question pertaining to additional medications or supplements, with only 18 owners confidently choosing to administer them at home, with an administration rate of 69.2%. Seven out of 40 (17.5%) dogs received adjuvant chemotherapy after complete incisional healing. Four of these patients received single-agent doxorubicin, while 3 received both doxorubicin and temozolomide. Of surgical patients surviving to discharge, 3 of 36 (8.3%) experienced owner-reported wound complications.
Median initial visit cost was $1,323.80 (range, $589.02 to $3,613.25). Median cost of surgery was $4,640.75 (range, $3,967.94 to $5,706.37). Median cost of nonsurgical visits was $716.50 (range, $518.28 to $1,497.20). Median cost to owners of patients that died or were euthanized in < 24 hours was $681.60 (range, $508.25 to $1,275.16). Median cost of palliative treatment was $1,143.02 (range, $600.73 to $2,199.48). Palliative treatment was more expensive than euthanasia (P = .021), and surgery was more expensive than palliative treatment (P < .001).
Influences on initial decision-making
Owners ranked QOL as the most important factor influencing their decision, followed by either risk of cancer or time remaining with their pet (ie, prognosis), then degree of invasiveness of treatment, or age (Table 1). Least important factors were presence of concurrent diseases and finances. Age did not influence decision to pursue surgery (P = 0.401).
Summary of owner-reported (n = 132) influences on treatment decision at the time of emergent presentation of dogs with nontraumatic hemoabdomen.
| Owner ranking | Quality of lifea | Risk of cancerb | Remaining time with petb | Invasive nature of treatmentc | Age of dogc | Concurrent diseased | Financesd |
|---|---|---|---|---|---|---|---|
| 1 | 3 (2.3%) | 24 (18.2%) | 20 (15.2%) | 29 (22.0%) | 33 (25.0%) | 60 (45.5%) | 52 (39.4%) |
| 2 | 1 (0.8%) | 7 (5.3%) | 13 (9.9%) | 16 (12.1%) | 22 (16.7%) | 18 (13.6%) | 30 (22.7%) |
| 3 | 6 (4.6%) | 26 (19.7%) | 25 (18.9%) | 39 (29.6%) | 39 (29.6%) | 25 (18.9%) | 29 (21.0%) |
| 4 | 42 (31.8%) | 47 (35.6%) | 43 (32.6%) | 27 (20.5%) | 23 (17.4%) | 21 (15.9%) | 18 (13.6%) |
| 5 | 80 (60.6%) | 28 (21.2%) | 31 (23.5%) | 21 (15.9%) | 15 (11.4%) | 8 (6.1%) | 3 (2.3%) |
| ≥ 4 | 122 (92.4%) | 75 (56.8%) | 74 (56.1%) | 48 (36.4%) | 38 (24.2%) | 29 (22.0%) | 21 (15.9%) |
| Median | 5 | 4 | 4 | 3 | 3 | 2 | 2 |
| IQR | 4–5 | 3–4 | 2.5–4 | 2–4 | 1.5–4 | 1–3 | 1–3 |
Owners were asked to rank the influence of the above factors on their treatment decision on a 5-point Likert scale (1 = none at all, 2 = very little, 3 = somewhat, 4 = heavily, 5 = entirely). a–dComparison of medians yielded a P value < .05 with differing superscripts.
Clinician experience (intern with < 1 year of experience vs clinician with > 1 year of experience) did not influence decision to pursue treatment (palliative care or surgery vs euthanasia; P = .305). Patients that had preoperative imaging suggesting hemorrhage from the spleen were significantly more likely to undergo surgery than those that had hemorrhage suspected from another source (OR, 29.333; CI, 7.6169 to 112.3597; P < .001). Patients with either presumptive (P = .001) or confident (P < .001) imaging metastasis were less likely to undergo surgery. Patients with confident metastasis (P < .001) were less likely to undergo any treatment option (palliative care or surgery). Patients that were euthanized at presentation had higher lactate values (n = 49; median, 6.54 [range, 4.03 to 9.26]) compared with those undergoing any treatment (n = 55; median, 2.9 [range, 1.99 to 3.99]; P < .001). Patients that underwent surgery had lower lactate values (n = 37; median, 2.99 [range, 2.22 to 3.79]) compared with those patients undergoing either palliative treatment or euthanasia (n = 67; median, 4.73 [range, 2.9 to 8.1]; P = .002).
Assessing treatment impacts
The majority of owners were satisfied with their treatment decision and would make the same decision again with the knowledge of the outcome obtained (Table 2). Owners were more likely to be satisfied (score ≥ 4 of 5) than indifferent or unsatisfied (score < 4 of 5) when choosing surgery over either euthanasia or palliative care (P = .039). Significance was lost when comparing any treatment versus euthanasia (P = .122). No meaningful differences were identified when evaluating treatment selected and rate of second guessing, likelihood of choosing the same treatment option again, or likelihood of choosing a different specific treatment option (P > .05). With regard to the choice between surgery or palliative care, there was no difference in frequency of owners reporting a worse-than-expected outcome (P = .541).
Summary of owner-reported treatment choice satisfaction.
| Survey question | Response choice | No. of responses |
|---|---|---|
| How satisfied were you with the overall outcome of your treatment choice? | Extremely unsatisfied | 15/132 (11.4%) |
| Somewhat unsatisfied | 9/132 (6.8%) | |
| Neither satisfied nor unsatisfied | 30/132 (22.7%) | |
| Somewhat satisfied | 21/132 (15.9%) | |
| Extremely satisfied | 57/132 (43.2%) | |
| Did you have second thoughts about the treatment decision you made? | Yes (second guessed) | 26/132 (19.7%) |
| No (did not second guess) | 99/132 (75.0%) | |
| Unsure | 7/132 (5.3%) | |
| If you were faced with this choice again, would you make the same treatment decision? | Yes (same decision) | 108/132 (81.8%) |
| No (different decision) | 11/132 (8.3%) | |
| Unsure | 13/132 (9.9%) | |
| If you would make a different treatment decision or were unsure, which treatment course would you consider instead? | Euthanasia | 5/24 (20.8%) |
| Palliative care | 5/24 (20.8%) | |
| Surgery | 5/24 (20.8%) | |
| Unsure | 9/24 (37.5%) | |
| How did your pet’s outcome compare with your expectation? | Better than expected | 27/60 (45%) |
| Same as expected | 17/60 (28.3%) | |
| Worse than expected | 16/60 (26.7%) |
Four-point Likert scale scores for QOL, appetite, comfort, and activity were evaluated for all patients prior to NTH onset, at NTH presentation, and again after treatment for those surviving to discharge, as well as for surgery and palliative care subgroups (Tables 3 and 4). No significant differences were found for any score between treatment categories either before or during the NTH event. However, QOL, appetite, comfort, and activity scores were all significantly higher afterwards for patients that had undergone surgery as compared with those that had undergone palliative care. Similar comparisons were performed for subgroup of patients that underwent surgery and were diagnosed with malignant disease, and significance was maintained for average postoperative QOL (P = .042), best QOL after (P = .043), and activity scores (P = .032). The presence of either presumptive or confident metastasis had no influence on any postoperative QOL parameters (P > .05).
Summary of owner-reported quality-of-life (QOL) scores on average, at the pet’s best, and at the pet’s worst prior to onset of nontraumatic hemoabdomen; on average while suffering from nontraumatic hemoabdomen; and on average, at the pet’s best, and at the pet’s worst following treatment for those that survived to discharge. Results are categorized into scores for all patients combined and then into the subgroups of patients that received palliative care and surgery. All: n = 132 before and during, 56 after. Palliative care: n = 22 before and during, 20 after. Surgery: n = 40 before and during, 36 after.
| QOL score category | QOL score for all patients | QOL score for palliative care patients | QOL score for surgery patients | P value |
|---|---|---|---|---|
| On average before NTH | 4 (3–4) | 4 (3–4) | 4 (3–4) | .689 |
| At the pet’s best before NTH | 4 (3–4) | 4 (3–4) | 4 (3.5–4) | .597 |
| At the pet’s worst before NTH | 3 (2–4) | 3 (2–4) | 3 (2–4) | .547 |
| On average during NTH event | 1 (1–2) | 2 (1–2) | 1 (1–2) | .268 |
| On average after treatment | 3 (2–3.5) | 2 (2–3) | 3 (3–4) | .007 |
| At the pet’s best after treatment | 3 (2–4) | 3 (1.5–3) | 4 (3–4) | .022 |
| At the pet’s worst after treatment | 2 (1–3) | 1 (1–2) | 3 (1–3.5) | .003 |
Owners scored their pet’s QOL on a 4-point Likert scale (1 = poor, 2 = fair, 3 = good, 4 = excellent). P value is a comparison of QOL scores between palliative care versus surgery patients. Nonnormally distributed data are presented as median (range).
NTH = Nontraumatic hemoabdomen.
Summary of owner-reported appetite, comfort, and activity scores for patients on average prior to onset of nontraumatic hemoabdomen, while suffering from nontraumatic hemoabdomen, and after treatment for those that survived to discharge. Results are categorized into scores for all patients combined and then into the subgroups of patients that received palliative care and surgery. All: n = 132 before and during, 56 after. Palliative care: n = 2 before and during, 20 after. Surgery: n = 40 before and during, 36 after.
| Score category | Score for all patients | Score for palliative care patients | Score for surgical patients | P value |
|---|---|---|---|---|
| Appetite | ||||
| On average before NTH | 4 (3–4) | 3.5 (3–4) | 4 (3–4) | .124 |
| On average during NTH event | 1 (1–2) | 2 (1–3) | 1 (1–3) | .093 |
| On average after treatment | 3 (2–4) | 2 (1.5–3) | 3.5 (2.5–4) | .011 |
| Comfort | ||||
| On average before NTH | 4 (3–4) | 3.5 (3–4) | 4 (3–4) | .353 |
| On average during NTH event | 1 (1–2) | 2 (1–2) | 1 (1–2) | .325 |
| On average after treatment | 3 (2–4) | 2 (1.5–3) | 3 (2.5–4) | .036 |
| Activity level | ||||
| On average before NTH | 4 (3–4) | 4 (3–4) | 4 (3–4) | .734 |
| On average during NTH event | 1 (1–2) | 1.5 (1–2) | 1 (1–2) | .446 |
| On average after treatment | 3 (2–4) | 2 (1.5–3) | 3 (3–4) | .007 |
See Table 3 for key.
Survival analysis
Of the 58 patients that survived to discharge, 38 (65.5%) were euthanized, 9 (15.5%) died naturally at home, and 11 (17.2%) remained alive at the time of survey dispersion. One patient did not have a reliable date of death. Reasons for euthanasia were often multiple in number and varied widely (Table 5). The reason for euthanasia for patients that were discharged was confidently not related to the initial diagnosis of NTH in 9 of 38 (23.7%) cases, with 1 patient being treated palliatively and 8 patients treated surgically. The median survival time (MST) was 1 day when considering all patients (n = 131; 95% CI, 1 to 4 days). MST was 213 days (n = 40; 95% CI, 74 to 826 days) for patients that underwent surgery. MST of patients undergoing palliative care was 39 days (n = 21; 95% CI, 8 to 549 days). Patients undergoing surgical treatment had longer survival times compared with patients undergoing palliative care (hazard ratio, 0.505; 95% CI, 0.256 to 0.996; P = .049). However, survival benefit of surgery was lost when considering patients with malignant histopathology (n = 30; MST, 81 days; 95% CI, 50 to 235 days) versus patients undergoing palliative care (hazard ratio, 0.697; 95% CI, 0.350 to 1.389; P = .305). On univariable survival analysis, suspected or confident presence of metastasis prior to histopathological diagnosis, as well as identification of malignant histopathology, negatively influenced survival (Table 6).
Summary of reasons why owners elected to euthanize their pet after initial hospital discharge following surgical or medical treatment for nontraumatic hemoabdomen. Owners could select more than one reason for euthanasia if applicable.
| Reason for euthanasia | No. of responses |
|---|---|
| Stopped eating/drinking | 16/38 (42.1%) |
| Slowed down, lethargic | 13/38 (34.2%) |
| Difficulty breathing | 13/38 (34.2%) |
| Recurrent hemoabdomen | 8/38 (21.1%) |
| Presence of metastatic disease | 12/38 (31.6%) |
| Comorbidity | 5/38 (13.2%) |
| Advanced age | 6/38 (15.8%) |
| Other | 18/38 (47.4%) |
Univariable survival analysis of patients with nontraumatic hemoabdomen.
| Possible prognostic factor | n | Presence (95% CI; d) | Absence (95% CI; d) | Hazard ratio | CI | P Value |
|---|---|---|---|---|---|---|
| Surgery | 61 | 213 (74–826) | 39 (8–549) | 0.51 | 0.26–1.00 | .049 |
| Potential imaging metastasis | 49 | 56 (28–155) | 1243 (74–?) | 4.13 | 1.67–10.20 | .002 |
| Confident imaging metastasis | 49 | 31 (1–64) | 654 (81–1243) | 4.58 | 2.10–9.98 | < .001 |
| Potential surgical metastasis | 40 | 31 (1–56) | 654 (140–1243) | 4.23 | 1.82–9.84 | .001 |
| Confident surgical metastasis | 40 | 31 (1–?) | 304 (99–922) | 2.49 | 1.04–5.94 | .040 |
| Malignant histopathology | 40 | 81 (50–235) | 922 (213–?) | 3.84 | 1.28–11.56 | .017 |
| Mitotic rate (mitoses/10 hpf) | 29 | — | — | 1.03 | 1.01–1.06 | .018 |
| Mitotic score 3 (vs 0–2) | 24 | 4 (3–?) | 113 (50–?) | 7.79 | 1.87–32.42 | .005 |
| PCV (%) | 61 | — | — | 0.94 | 0.90–0.98 | .003 |
| Hct (%) | 58 | — | — | 0.93 | 0.89–0.97 | .001 |
| Platelet count (thousand/UL) | 37 | — | — | 0.99 | 0.99–1.00 | .008 |
| Lactate (mmol/L) | 55 | — | — | 0.96 | 0.83–1.11 | .612 |
| TFAST fluid | 34 | 3 (?–?) | 213 (50–922) | 12.25 | 1.08–138.86 | .043 |
| TFAST B-lines | 24 | 140 (99–?) | 654 (31–1243) | 1.30 | 0.36–4.75 | .691 |
| Blood transfusion | 61 | 155 (50–826) | 94 (33–549) | 0.95 | 0.51–1.80 | .885 |
| Age (y) | 61 | — | — | 1.19 | 0.56–2.52 | .649 |
| Weight (kg) | 61 | — | — | 1.00 | 0.98–1.03 | .710 |
| Any supplements administered at home | 42 | — | — | 1.33 | 0.59–3.00 | .487 |
| Time until surgery (h) | 39 | — | — | 1.01 | 0.98–1.05 | .421 |
| Source of hemorrhage from spleen | 40 | 213 (74–922) | 826 (1–?) | 1.13 | 0.26–4.90 | .868 |
| Seen by an intern | 40 | 213 (81–922) | 155 (19–?) | 0.81 | 0.37–1.77 | .601 |
Nonnormally distributed data are presented as median (range).
— = Presence/absence data not available for continuous variable. ? = CI unable to be obtained due to small sample size. TFAST = Thoracic focused assessment with sonography for trauma.
Discussion
To the authors’ knowledge, this is the first large retrospective study to evaluate factors impacting owner decision-making regarding treatment of dogs presenting emergently with NTH. Additionally, this study assessed owner satisfaction with a chosen treatment and perceived QOL before, during, and after treatment. The 3 treatment options evaluated were euthanasia, palliative care, and surgery. The majority of patients were euthanized, consistent with previous studies2,3 on dogs with NTH. Patients more commonly received palliative care or were euthanized if initial workup revealed elevated lactate or imaging was highly suggestive of metastasis. Patients undergoing surgery had significantly increased MST with greater perceived QOL and owner satisfaction compared with both palliative care and euthanasia. Although this survival benefit of surgery was lost when patients without malignant disease were excluded, several QOL scores remained improved.
Owners reported QOL as the most important influence on decision-making, heavily or entirely impacting the treatment choice of > 90% of owners. Owners of dogs that underwent surgery reported significantly better QOL scores following treatment compared with those receiving palliative care, including increased appetite, comfort, and activity. After selecting for the subgroup of dogs with surgery and malignant disease, significantly higher activity and QOL scores on average and at the pet’s best were maintained as compared with scores for patients receiving palliative care. Given our findings, the apparent QOL advantage reported by owners of animals receiving surgery should be included in initial counseling, although these results are prone to recall and observer biases in this nonrandomized, nonblinded retrospective study design. While multiple factors, including satisfaction and survival, influence owners deciding between palliative care and surgery, our findings suggest that if QOL is identified as the main guiding concern in an owner’s selection, surgery should be highly considered.
Following QOL, owners considered risk of cancer and time remaining with their pet as the next most influential factors on treatment decision. Some prognostic indicators and diagnostic criteria, such as the hemangiosarcoma likelihood prediction (HeLP) score, have been recognized to prioritize neoplastic disease in patients with NTH.5–11,17 However, until a definitive diagnosis is able to be made prior to histopathology, discussing the risk of common cancers and their associated prognoses in NTH cases is prudent. Many clinicians use the double two-thirds rule, which states two-thirds of splenic masses will be malignant and two-thirds of those will be hemangiosarcoma, to counsel owners of pets presenting with NTH10,18; however, a recent study19 that evaluated the validity of this rule in dogs with NTH from a ruptured splenic mass found more dogs were diagnosed with neoplasia (73.0%) and hemangiosarcoma specifically (87.3%) than the double two-thirds rule indicates. Our data support this, with 75.6% of cases with histopathology classified as malignant and 83.9% of those malignant cases attributed to hemangiosarcoma.19 Owners should therefore be educated on the prevalence of malignant diagnoses and splenic hemangiosarcoma in particular, as these cases carry a poor prognosis, with an MST of < 3 months without adjuvant chemotherapy.18–22 Specifically, patients with malignant histopathology had an MST of 81 days in this study.
Considering the reported influence of the risk of cancer on decision-making, caution must be exercised before recommending surgery in all cases. The most negative responses to the survey were from owners who felt they were not adequately advised on prognosis prior to pursuing surgery. When asked to describe why they would not make the same treatment decision again, these owners said, “poor prognosis for recovery after surgery … was not appropriately advised by medical staff, including the surgeon, prior to the surgery” and “the doctor gave us more hope than what should have been presented … had I been given a more honest answer about prognosis, I would have opted for humane euthanasia.” These responses underscore the need for standardized NTH discussions to improve client comprehension of treatment options and potential outcomes.
Though a definitive diagnosis prior to tissue acquisition is not currently feasible, screening for metastasis through preoperative imaging may improve confidence in a malignant pathology and may be useful for an owner’s decision-making process. Our findings highlight that high suspicion of metastasis on imaging was negatively associated with survival, making surgery a less advantageous treatment approach for owners most concerned with survivability of their pet. Out-of-hours staging may be warranted for owners who would euthanize their pet if metastasis were identified. Diagnostic modalities with higher sensitivity for metastasis may be beneficial for owners who are concerned about patient longevity. Studies show CT may be able to detect early pulmonary and abdominal metastasis and discriminate between benign and malignant splenic lesions.20,23–25 These results highlight the need for nuanced discussions and appropriate preoperative staging prior to hospital admission and treatment.
While QOL and malignant potential were reportedly the most important factors influencing owner decision-making, other factors such as finances, comorbidities, and pet age displayed modest importance. These findings demonstrate that an individualized approach to owner counseling is necessary to solicit goals of care and priorities. Although not a discrete comorbidity, patients with elevated lactate on presentation were less likely to undergo any treatment, as well as surgery specifically; however, if taken to surgery, patients with elevated lactate preoperatively were not negatively impacted with regard to long-term survival. This association may demonstrate a potential bias in how providers educate owners, and eventually treat their patients, when a dog presents with elevated lactate in the context of NTH.
While a large majority of our sample reported confidence in their treatment choice and satisfaction with the related outcome, a smaller proportion of our respondents, about 20%, definitively second-guessed their treatment choice, and over 5% were unsure if they had second thoughts. Additionally, about 20% of owners were either extremely or somewhat unsatisfied with the overall outcome of their treatment choice, while over 25% of owners reported a worse-than-expected patient outcome. These findings were not unique to any of the 3 treatment categories. Given the small proportion of our sample reporting these trends, we hope our study provides information to assist in informed decision-making of owners to promote confidence in treatment choices. To further mitigate owner dissatisfaction and second-guessing, incorporation of social workers into veterinary treatment teams has emerged as a promising strategy to aid in client counseling, education, and follow-up communication.26,27
The present study had several limitations. Owners responded to a retrospective survey sent by email, and data relied on owners’ ability to recall the factors impacting their treatment decision and satisfaction and emotions regarding that choice. Hence, the survey was prone to recall bias (follow-up times ranged from < 1 year to > 7 years) and nonrespondent bias. Further, the nonrandomized, nonblinded study design invoked a risk of observer bias, as owners who elected for surgery might have been more inclined to retroactively overestimate their pet’s QOL. Moreover, a validated veterinary QOL scoring system was not established in our study; therefore, owners were asked to rank their pet’s QOL on a 4-point Likert scale (1 = poor, 2 = fair, 3 = good, 4 = excellent). Differing interpretations of this subjective metric may have led to unreliability in QOL scores. Finally, through preoperative staging, a healthier group of individuals may have been selected for surgery, thereby skewing survival times.
We report here the findings of the first large retrospective study of factors impacting owner decision-making regarding treatment of dogs presenting emergently with NTH, specifically assessing owner satisfaction and perceived QOL. Owners reported QOL as the most important influence on decision-making for NTH. Patients that underwent surgery had significantly increased MSTs with greater perceived QOL and owner satisfaction compared with patients that underwent palliative care and euthanasia, suggesting that if these factors are identified as most important to a client, surgery should be considered. Further, owners considered risk of cancer and time remaining with their pet as the next most influential factors impacting their treatment choice. We found that presumptive or confident diagnosis of metastasis decreased survival times for all patients and that the survival benefit of surgery was lost when only those patients with confirmed malignant histopathology were analyzed. Therefore, we highly recommend preoperative staging and screening for metastasis, especially for owners who personally view malignant disease as a contraindication for treatment. Regardless of treatment choice, about 25% of our respondents second-guessed their decision, which highlights the need for further educational and emotional support of clients during this difficult decision. Ultimately, our findings strongly emphasize the need for standardization of NTH discussions of treatment options and potential outcomes, guided by factors that are of greatest importance to an individual pet owner, to effectively and efficiently guide treatment of patients with this common presentation.
Supplementary Materials
Supplementary materials are posted online at the journal website: avmajournals.avma.org
Acknowledgments
No third-party funding or support was received in connection with this study or the writing or publication of the manuscript. The authors have nothing to declare.
References
- 1.↑
Brockman DJ, Mongil CM, Aronson LR, Brown DC. A practical approach to hemoperitoneum in the dog and cat. Vet Clin North Am Small Anim Pract. 2000;30(3):657–668. doi:10.1016/s0195-5616(00)50044-8
- 2.↑
Pintar J, Breitschwerdt EB, Hardie EM, Spaulding KA. Acute nontraumatic hemoabdomen in the dog: a retrospective analysis of 39 cases (1987-2001). J Am Anim Hosp Assoc. 2003;39(6):518–522. doi:10.5326/0390518
- 3.↑
Aronsohn MG, Dubiel B, Roberts B, Powers BE. Prognosis for acute nontraumatic hemoperitoneum in the dog: a retrospective analysis of 60 cases (2003-2006). J Am Anim Hosp Assoc. 2009;45(2):72–77. doi:10.5326/0450072
- 4.
Fleming J, Giuffrida MA, Runge JJ, et al. Anatomic site and etiology of hemorrhage in small versus large dogs with spontaneous hemoperitoneum. Vet Surg. 2018;47(8):1031–1038. doi:10.1111/vsu.12963
- 5.↑
Lux CN, Culp WTN, Mayhew PD, Tong K, Rebhun RB, Kass PH. Perioperative outcome in dogs with hemoperitoneum: 83 cases (2005-2010). J Am Vet Med Assoc. 2013;242(10):1385–1391. doi:10.2460/javma.242.10.1385
- 6.↑
Millar SL, Curley TL, Monnet EL, Zersen KM. Premature death in dogs with nontraumatic hemoabdomen and splenectomy with benign histopathologic findings. J Am Vet Med Assoc. 2021;260(S1):S9–S14. doi:10.2460/javma.21.01.0033
- 7.
Schick AR, Hayes GM, Singh A, Mathews KG, Higginbotham ML, Sherwood JM. Development and validation of a hemangiosarcoma likelihood prediction model in dogs presenting with spontaneous hemoabdomen: the HeLP score. J Vet Emerg Crit Care (San Antonio). 2019;29(3):239–245. doi:10.1111/vec.12838
- 8.
Masyr AR, Rendahl AK, Winter AL, Borgatti A, Modiano JF. Retrospective evaluation of thrombocytopenia and tumor stage as prognostic indicators in dogs with splenic hemangiosarcoma. J Am Vet Med Assoc. 2021;258(6):630–637. doi:10.2460/javma.258.6.630
- 9.
Prymak C, McKee LJ, Goldschmidt MH, Glickman LT. Epidemiologic, clinical, pathologic, and prognostic characteristics of splenic hemangiosarcoma and splenic hematoma in dogs: 217 cases (1985). J Am Vet Med Assoc. 1988;193(6):706–712.
- 10.↑
Johnson KA, Powers BE, Withrow SJ, Sheetz MJ, Curtis CR, Wrigley RH. Splenomegaly in dogs. Predictors of neoplasia and survival after splenectomy. J Vet Intern Med. 1989;3(3):160–166. doi:10.1111/j.1939-1676.1989.tb03092.x
- 11.↑
Hammond TN, Pesillo-Crosby SA. Prevalence of hemangiosarcoma in anemic dogs with a splenic mass and hemoperitoneum requiring a transfusion: 71 cases (2003-2005). J Am Vet Med Assoc. 2008;232(4):553–558. doi:10.2460/javma.232.4.553
- 12.↑
Crawford AH, Tivers MS, Adamantos SE, Crawford AH. Owner assessment of dogs’ quality of life following treatment of neoplastic haemoperitoneum. Vet Rec. 2012;170(22):566. doi:10.1136/vr.100595
- 13.↑
Molitoris A, Pfaff A, Cudney S, de Laforcade A. Early career clinicians euthanize more dogs with nontraumatic hemoabdomen but not gastric dilatation and volvulus than more experienced clinicians. J Am Vet Med Assoc. 2022;260(12):1514–1517. doi:10.2460/javma.22.05.0198
- 14.↑
Dickerson VM, Coleman KD, Ogawa M, et al. Outcomes of dogs undergoing limb amputation, owner satisfaction with limb amputation procedures, and owner perceptions regarding postsurgical adaptation: 64 cases (2005-2012). J Am Vet Med Assoc. 2015;247(7):786–792. doi:10.2460/javma.247.7.786
- 15.↑
Tzannes S, Hammond MF, Murphy S, Sparkes A, Blackwood L. Owners ‘perception of their cats’ quality of life during COP chemotherapy for lymphoma. J Feline Med Surg. 2008;10(1):73–81. doi:10.1016/j.jfms.2007.05.008
- 17.↑
Burgess KE, Price LL, King R, et al. Development and validation of a multivariable model and online decision-support calculator to aid in preoperative discrimination of benign from malignant splenic masses in dogs. J Am Vet Med Assoc. 2021;258(12):1362–1371. doi:10.2460/javma.258.12.1362
- 18.↑
Spangler WL, Kass PH. Pathologic factors affecting postsplenectomy survival in dogs. J Vet Intern Med. 1997;11(3):166–171. doi:10.1111/j.1939-1676.1997.tb00085.x
- 19.↑
Schick AR, Grimes JA. Evaluation of the validity of the double two-thirds rule for diagnosing hemangiosarcoma in dogs with nontraumatic hemoperitoneum due to a ruptured splenic mass: a systematic review. J Am Vet Med Assoc. 2023;261(1):69–73. doi:10.2460/javma.22.08.0389
- 20.↑
Mullin C, Clifford CA. Miscellaneous tumors. In: Vail DM, Thamm DH, Liptak JM, eds. Withrow and MacEwen’s Small Animal Clinical Oncology. 6th ed. Elsevier; 2020:773–810.
- 21.
Wendelburg KM, Price LL, Burgess KE, Lyons JA, Lew FH, Berg J. Survival time of dogs with splenic hemangiosarcoma treated by splenectomy with or without adjuvant chemotherapy: 208 cases (2001-2012). J Am Vet Med Assoc. 2015;247(4):393–403. doi:10.2460/javma.247.4.393
- 22.↑
Wood CA, Moore AS, Gliatto JM, Ablin LA, Berg RJ, Rand WM. Prognosis for dogs with stage I or II splenic hemangiosarcoma treated by splenectomy alone: 32 cases (1991-1993). J Am Anim Hosp Assoc. 1998;34(5):417–421. doi:10.5326/15473317-34-5-417
- 23.↑
Fife WD, Samii VF, Drost WT, Mattoon JS, Hoshaw-Woodard S. Comparison between malignant and nonmalignant splenic masses in dogs using contrast-enhanced computed tomography. Vet Radiol Ultrasound. 2004;45(4):289–297. doi:10.1111/j.1740-8261.2004.04054.x
- 24.
Armbrust LJ, Biller DS, Bamford A, Chun R, Garrett LD, Sanderson MW. Comparison of three-view thoracic radiography and computed tomography for detection of pulmonary nodules in dogs with neoplasia. J Am Vet Med Assoc. 2012;240(9):1088–1094. doi:10.2460/javma.240.9.1088
- 25.↑
Borgatti A, Winter AL, Stuebner K, et al. Evaluation of 18-F-fluoro-2-deoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) as a staging and monitoring tool for dogs with stage-2 splenic hemangiosarcoma—a pilot study. PLoS One. 2017;12(2):e0172651. doi:10.1371/journal.pone.0172651
- 26.↑
Siess S, Marziliano A, Sarma EA, Sikorski LE, Moyer A. Why psychology matters in veterinary medicine. Top Companion Anim Med. 2015;30(2):43–47. doi:10.1053/j.tcam.2015.05.001
- 27.↑
Netting FE, Wilson CC, New JC. The human-animal bond: implications for practice. Soc Work. 1987;32(1):60–64. doi:10.1093/sw/32.1.60
