Time course of clinical signs and mortality in dogs with severe perioperative acute kidney injury: A scoping review

Abstract

Perioperative acute kidney injury (AKI) is a potential cause of anaesthetic mortality in dogs. The time delay between anaesthetic recovery, onset of clinical signs and any subsequent mortality may result in under‐recognition of this complication. This review aimed to explore the literature reporting dogs with severe AKI after general anaesthesia and surgery. Firstly, to determine the time course between anaesthesia recovery and onset of clinical signs, and between recovery and any mortality. Secondly, to identify the common clinical signs and signalment of dogs with perioperative AKI. PubMed and CAB abstracts data bases using the terms “(acute kidney injury OR acute renal failure) AND dog AND (anaesthesia OR surgery)”; and ResearchRabbit were searched. Peer reviewed publications in English describing dogs that developed AKI with overt clinical signs after anaesthesia were included. Number of postoperative days until onset of clinical signs and death; along with signalment and the reported clinical signs leading to AKI diagnosis were extracted. Nine publications describing a total of 31 dogs were included in the review. Clinical signs were typically first seen 2–4 days postoperatively (range 1–14). Death/euthanasia occurred in 5 dogs; between 3 and 60 days postoperatively. Persistent renal dysfunction occurred in 4 survivors. The most common clinical signs were anorexia, lethargy, polyuria/polydipsia and vomiting. Female and larger breed dogs especially Labradors and Golden Retrievers were overrepresented. Knowledge of this time course may improve postoperative monitoring and recognition of perioperative AKI in dogs.

Abbreviations

AKI

Acute Kidney Injury

BW

Body Weight

CI

Confidence Interval

COX‐2

cyclooxygenase‐2

IV

Intravenous

MAKE

Major Adverse Kidney Event

PEO

Population Exposure Outcome

PRISMA‐ScR

Preferred Reporting Items for Systematic reviews and Meta‐Analyses extension for Scoping Reviews

RRT

Renal Replacement Therapy

UPC

Urine Protein to Creatinine ratio

Acute kidney injury (AKI) is a cause of anaesthetic related death in dogs. ¹ , ² Anaesthetic mortality studies have reported renal disease as the cause of death in 1%–14% of deaths. ¹ , ² It is likely that this is a substantial underestimate of the true incidence. As the follow‐up period of veterinary anaesthetic mortality studies is generally short, only animals that rapidly developed AKI and died within the first hours to days postoperatively would be included. Secondly, dogs that developed AKI but survived (at least for a few days) would not be included. Thirdly, mortality studies may exclude deaths that could be attributed to causes other than anaesthesia. ³ Perioperative AKI is a multifactorial complication ⁴ that may not meet the case definition of mortality studies. For example, if exposure to a nephrotoxin such as iodinated contrast ⁵ or hypovolaemia due to surgical blood loss occurred, the case might be excluded. Delay in the onset of clinical signs may also make it difficult for veterinarians to identify individual cases of AKI as a perioperative complication, further contributing to the unrecognition of the problem.

In humans, AKI is a well‐known complication of surgery and anaesthesia. ⁴ , ⁶ Large studies in human surgical populations have shown that even sub‐clinical AKI is associated with an increased risk of further in‐hospital complications, development of chronic kidney disease and increased all‐cause mortality. ⁷ To date, there have been no large studies evaluating perioperative AKI in dogs. Small studies in critically ill dogs undergoing referral surgery have identified incidences of up to 40% in these presumably high‐risk patients. ⁸ , ⁹ , ¹⁰ Combined with the findings from anaesthetic mortality studies, this supports the suggestion that AKI may be a similar problem in dogs as it is in human surgical patients.

Knowledge of the time course of perioperative AKI in dogs may be of value to both practicing veterinarians and for research evaluating this complication. Currently, the typical duration between anaesthetic recovery and onset of clinical signs or subsequent mortality is unknown. This information is essential for designing an appropriate study to determine the incidence of this condition in the wider canine surgical population. Improved postoperative monitoring in the days after anaesthesia has been identified as an important step for reducing canine perioperative morbidity and mortality. ³ , ¹¹ Knowledge of this time course would provide evidence for improving postoperative care, including appropriate duration for monitoring and thus enabling more rapid diagnosis if a patient presents with relevant clinical signs within this period. Secondly, identification of common clinical signs within this time frame will enable targeted instructions to dog owners for postdischarge monitoring to be provided, enabling timely readmission and treatment.

The objective of this review was to search the literature reporting on dogs with severe AKI (i.e. with sufficient severity to lead to overt clinical signs) after general anaesthesia to address the following questions in this population exposure outcome (PEO) setting. Firstly, how long after general anaesthesia do clinical signs develop and any subsequent mortality occur? Secondly, in this PEO setting, what are the common presenting clinical signs and signalment?

Methods

Information sources and search strategy

The review was structured following the recommendations outlined in Preferred Reporting Items for Systematic reviews and Meta‐Analyses extension for Scoping Reviews (PRISMA‐ScR) Checklist. ¹² The PubMed and CAB abstracts data bases were searched for peer reviewed publications reporting on dogs with diagnosed perioperative AKI. As the terminology relating to AKI has evolved over time the historical term “acute renal failure” was also included. The following search terms were used: “(acute renal failure OR acute kidney injury) AND dog AND (anaesthesia OR surgery).” Searches were limited to publications in English, both data bases were searched from inception (1900 and 1910 respectively) to the final date of final search (completed 23^rd December 2024). Only publications that had undergone peer review were considered eligible for inclusion.

Titles and abstracts from the search of each data base were assessed to identify potentially eligible publications. After excluding duplicates, full‐text articles of publications identified from the title /abstract search were obtained and assessed for inclusion. The reference lists of included articles were also searched for additional potentially eligible publications. A further search of related literature was conducted using the artificial intelligence literature mapping tool “ResearchRabbit” (https://researchrabbitapp.com). To conduct this search, the two publications with the largest number of eligible cases identified from the data base searches ¹³ , ¹⁴ were seeded into the tool, and the resulting literature maps for earlier, later and similar work were examined. Additional publications were identified for title and abstract assessment. The final literature map was generated on 24 December 2024.

Selection of sources of evidence

Publications were considered if they described one or more dogs that had undergone general anaesthesia and had a subsequent new diagnosis of severe AKI, substantial worsening of existing kidney disease (acute on chronic kidney disease) or had changes in serum creatinine measurements consistent with a diagnosis of AKI using current guidelines with overt clinical signs. ¹⁵ Publications were also included if the diagnosis of new onset kidney disease was made at postmortem via histopathology consistent with severe renal disease. For the purposes of this review, severe AKI was considered to be AKI with overt clinical signs and sufficient severity to warrant treatment and/or euthanasia. Publications documenting AKI via novel biomarkers alone were excluded. Eligible publications were included if they provided details of the time to onset of clinical signs after recovery from anaesthesia, or time to death/euthanasia after anaesthesia, or description of the clinical signs attributable to AKI. Both clinical and laboratory settings were included if the above information was reported.

Data extraction and synthesis of results

The type of publication along with the information addressing the research questions was extracted from each eligible publication and entered into a spreadsheet to generate summary tables (Microsoft excel, Microsoft Corporation, CA, USA). Year and type of publication (case report, case series, case–control study, cohort study or experimental) and number of animals described were recorded. Where available, individual animal details were extracted including age, breed, body weight (BW), sex and procedure. Evidence of any pre‐existing renal disease was noted along with criteria used to diagnose AKI. In addition, details of perioperative care were extracted as available including anaesthetic drugs; use of non‐steroidal anti‐inflammatory drugs (NSAIDS); use of blood pressure monitoring or evidence of intraoperative hypotension/hypovolaemia; and use of intravenous fluids. Where relevant and feasible, summary descriptive statistics were calculated with data reported as median (range); or number and (proportion %) of cases.

To address the aims of the review the following were identified:

• Number of hours or days after recovery from anaesthesia (defined as completion of the procedure) that clinical signs leading to the diagnosis of AKI were first observed.

• Number of hours or days after recovery from anaesthesia until death or euthanasia. Euthanasia was included if it was in response to the severity of renal disease, either due to declined treatment or failure to respond to treatment. Euthanasia due to pre‐existing or unrelated conditions was excluded from this analysis.

• New clinical signs that prompted investigation leading to AKI diagnosis.

• Use of renal replacement therapy (RRT)

• In survivors, if there was evidence of persistent renal disease after discharge from hospital.

Synthesis of results

Descriptive statistics were calculated from the above information including median (range) for time to onset of clinical signs and death/euthanasia. The proportions of cases with each clinical sign were calculated. The number and proportion of survivors with evidence of persistent kidney disease were calculated. The proportion of animals with Major Adverse Kidney Events (MAKE), a composite outcome measure consisting of death, need for RRT and persistent kidney dysfunction, was calculated. ¹⁶ The significance of the proportion of female dogs was assessed using online statistical software, assuming a null hypothesis that the proportion of female to male dogs would be equal (i.e. 50% female) (MedCalc Software Ltd. Test for one proportion calculator https://www.medcalc.org/calc/test_one_proportion.php Version 23.1.1; accessed January 7, 2025). A P value less than 0.05 was considered significant. Proportions of other variables were not evaluated as there was no reliable population distribution to compare with.

Results

Data base searches retrieved 350 English language publications from PubMed and 60 from CAB abstracts. Title and abstract screening identified 26 and 9 potentially eligible publications from PubMed and CAB respectively, all of which were from peer‐reviewed journals (Figure 1). After removal of duplicates and full text review, five eligible publications were identified. One further publication was excluded as it was identified as a preliminary report ¹⁷ of a subsequent publication. ¹⁴ Reference lists from included papers and literature maps generated by Research Rabbit identified a further five eligible publications after full text assessment. Thus, a total of nine eligible publications were included in the final review (Figure 1).

Figure 1.

PRISMA‐ScR flow diagram of literature search process.

The nine publications consisted of five single animal case reports, three case series and one retrospective case–control study (Table 1). These provided detailed descriptions of 31 dogs (21 female, 5 male and 5 sex unreported) diagnosed with perioperative AKI over a 34‐year period (1990–2024). The majority of publications were by authors from academic institutions; however, some of the cases reported had been referred from private practice after anaesthesia due to the onset of clinical signs. One case series was reported jointly by four private practices. ¹³ One case series described dogs that had been used in a veterinary school surgical training facility; minimal details were provided about the signalment of these 5 cases. ¹⁸ All other publications described clinical cases (n = 26).

Table 1.

Results from individual publications included in the final review of the time course of clinical signs in dogs with perioperative acute kidney injury. Included publications are presented in chronological order.

Reference	Year	Study type	Number of dogs	Female (n)	BW (kg)	Onset of clinical signs (days)	Time to death (days)	Mortality (n)	MAKE (n)
Mathews 18	1990	Case series	5	Not reported	Not reported	1–1.5	3	1	2
McNeil 20	1992	Case report	1	1	20	3	4	1	1
Elwood 21	1992	Case series	2	2	25 & 38	1–2	12	1	2
Pascoe 14	1996	Case–control study	7	4	21–60	2–4	7	1	1
Falkeno 24	2014	Case report	1	1	Not reported	1	N/A	0	0
Torrente 23	2019	Case report	1	1	23	2	N/A	0	1
Rogers‐Smith 13	2020	Case series	12	10	24–56	2–14	N/A	0	1
Ku 22	2023	Case report	1	1	3.5	1	N/A	0	0
Johnson 19	2024	Case report	1	1	17.3	1	60	1	1
Total	–	–	31 (26 sex reported)	21 (81%)	3.5–60	1–14	3–60	5 (16%)	9 (29%)

Body weight (BW) in kg, onset of clinical signs in days and time to death in days are reported as range for case series and observational studies and exact number for individual case reports. Mortality and Major Adverse Kidney Events (MAKE) are reported as number of dogs and % of cases.

Clinical signs were first noted a median of 2 (range 1–14) days after recovery from anaesthesia. The majority of cases (n = 18, 58%) were noted to develop signs between 2 and 4 days after anaesthesia (Table 1). Five dogs died or were euthanased due to the severity of renal disease (16% mortality). Death or euthanasia occurred a median of 7 (range 3–60) days after anaesthesia and surgery (Table 1). Treatment/supportive care with at least intravenous fluid therapy had been attempted in all fatal cases. One dog was treated with RRT ¹⁹ and 5 (16%) dogs developed persistent renal dysfunction. Thus, MAKE occurred in 9 (29%) of the reported cases.

The most common clinical signs reported when dogs represented after anaesthesia were anorexia (n = 23, 74%), lethargy/depression (n = 21, 68%), polyuria/polydipsia (n = 17, 55%) and vomiting (n = 12, 39%) (Table 2). Anorexia, vomiting, lethargy/depression and polyuria/polydipsia were reported in the majority of publications (Table 2). Oliguria was not reported. However, no report provided quantitative urine output measurements to confirm polyuria or definitively exclude oliguria. Serum creatinine was elevated consistent with AKI in 30 cases and not measured in one case. ²⁰ In the latter case, postmortem examination revealed a diagnosis of acute tubulointerstitial nephritis. ²⁰ Post‐mortem findings were available from two other fatal cases ¹⁸ , ²¹ and similarly reported tubulointerstitial nephritis.

Table 2.

Clinical signs reported in 31 dogs from 9 peer‐reviewed publications with severe perioperative acute kidney injury

Clinical sign	Number of dogs	% of cases	Number of publications reporting clinical sign
Vomiting	12	38.7%	6 14 , 18 , 19 , 20 , 21 , 22
Anorexia	23	74.2%	7 13 , 14 , 18 , 19 , 20 , 21 , 23
Diarrhoea	3	9.7%	2 14 , 19
Lethargy/depression	21	67.7%	7 13 , 14 , 18 , 19 , 21 , 22 , 23
Polyuria/polydipsia	17	54.8%	5 13 , 14 , 18 , 21 , 23
Anuria or oliguria	0	0%	0
Incontinence	1	3.2%	1 13
Dehydration	11	35.5%	4 13 , 14 , 21 , 24
Pyrexia	1	3.2%	1 13

Female dogs were significantly overrepresented (21 of 26 with reported sex, 81%) (95% Confidence Interval [CI]: 60.91% to 93.58%; P = 0.0016). Seventeen female dogs were entire at the time of anaesthesia and surgery and four were neutered. None were reported to be pregnant nor lactating. Of the males, one was entire and 4 were neutered. Age ranged from 6 months to 13 years, with the majority (n = 23 of 31, 74%) between 6 months and 6 years. Body weight was reported in 26 cases and ranged from 3.5 to 60 kg. For the majority of dogs (n = 24 of 26, 92%) BW was >20 kg, and the remaining two dogs weighed 17.3 and 3.5 kg. Of the 12 breeds reported, only 2 were represented more than once: Labrador retriever (n = 4, 13%) and Golden retriever (n = 3, 10%). Labradors were recorded in one study ¹³ and Golden retrievers in three. ¹³ , ²⁰ , ²¹ The most common surgical procedures were ovariohysterectomy (n = 10, 32%), orthopaedic surgery (n = 9, 29%) and laparoscopic ovariectomy (n = 3, 10%). When female desexing surgeries (ovariohysterectomy or laparoscopic ovariectomy) were excluded, female sex was no longer significantly more common (n = 8 of 13 dogs, 62%) (95% CI: 31.98% to 86.45%; P = 0.3869). However, the proportion of dogs with BW > 20 kg remained high (n = 17, 94%).

Clinical pathology was inconsistently reported. Blood urea nitrogen was elevated in all but one case ²² where it was measured (n = 15 cases of 16). Hyponatraemia (n = 8 of 10) ¹⁴ , ²² and hyperkalaemia (n = 4 of 9) ¹⁴ , ²² , ²³ were the most common electrolyte abnormalities but were not consistent. Hyperphosphataemia was only reported in 3 cases (n = 3 of 12). ¹³ Urinalysis data were reported for 22 cases. ¹³ , ¹⁴ , ²¹ , ²² , ²³ , ²⁴ Isosthenuria was a consistent finding (n = 18 of 22, 82%; or 58% of the total 31 cases). Marked increase (>>0.5) in urine protein to creatinine ratio (UPC) was reported in 6 of 10 cases. ¹³

Anaesthetic drugs varied considerably between reports and across dates of publication. Premedication most commonly included an opioid (n = 21, 75%) ¹³ , ¹⁴ , ¹⁸ , ²³ with or without either acepromazine (n = 15, 54%) ¹³ , ¹⁴ , ¹⁸ , ²⁰ or an alpha‐2 agonist (medetomidine or dexmedetomidine) (n = 12, 43%). ¹³ Anaesthesia was induced with either a barbiturate (n = 16, 52%), ¹⁴ , ¹⁸ , ²⁰ , ²¹ propofol (n = 13, 42%) ¹³ , ¹⁴ , ²² or alfaxalone (1 case) ¹³ and maintained with an inhalational anaesthetic in all but 1 case. ¹⁴ Inhalational agents included isoflurane (n = 15, 48%), ¹³ , ¹⁴ , ²² , ²³ halothane (n = 8, 26%) ¹⁴ , ²⁰ , ²¹ or methoxyflurane (n = 5, 16%). ¹⁸ Intravenous (IV) fluids were reported as administered perioperatively in 23 cases (74%) in 5 publications. ¹³ , ¹⁴ , ¹⁸ , ²² , ²³ Where given, intraoperative fluids were administered at a median rate of 5 (range 4–20) ml/kg/h. Most authors reported the use of a polyionic balanced crystalloid solution for fluid therapy, although 0.9% saline was used in one case. ²² Blood pressure monitoring was performed in 26 cases (84%) in 5 publications. ¹³ , ¹⁴ , ¹⁸ , ²² , ²³ A brief period of hypotension (mean arterial pressure < 60 mmHg) was reported in two cases in report ¹³ although the actual mean arterial pressures and the interval between measurements were not reported. Hypotension was not reported in any other study. However, definitions of hypotension were not provided. NSAIDS were administered in 19 cases (74%); these were meloxicam (n = 9), ¹³ flunixin (n = 8) ¹⁸ , ²⁰ , ²¹ and carprofen (n = 2). ¹³ , ²⁴ NSAIDs were given preoperatively in 5 cases, intraoperatively in 6 and at some point in recovery in 8. Other potential nephrotoxins used perioperatively included methoxyflurane in 5 cases from 1 report ¹⁸ and the antibiotics nafcillin (7 cases) ¹⁴ or sulphonamides (1 case). ²⁰

As one report ¹³ provided 12 of the 31 cases (39%) the distribution of sex, BW and procedure were considered without this publication. This was completed to determine if the same trends could be observed without the influence of the findings of this disproportionately large case series that may otherwise have distorted the results of the review. Of the remaining 19 cases, 11 of 15 cases (89%) had BW >20 kg. Ovariohysterectomy (n = 6, 32%) and orthopaedic surgery (n = 9, 47%) remained the most common procedures. Golden retrievers were the only breed reported more than once in these reports. Of the 14 cases with known sex, 11 (79%) were female (95% CI: 49.66% to 95.54%; P = 0.03).

Discussion

Clinical signs leading to a diagnosis of AKI were typically first noted between 2 and 4 days postoperatively in the dogs included in the reviewed publications. No clinical signs were noted earlier than 1 day and at the latest by 14 days. Death or euthanasia due to AKI occurred between 3 and 60 days after recovery from anaesthesia. Overall MAKE occurred in 29% of cases, indicating a relatively high morbidity and mortality burden with this perioperative complication. Further research into perioperative AKI in dogs is needed and should take this time course into consideration to obtain an accurate assessment of the prevalence of this complication.

The predominant clinical signs reported were anorexia, lethargy and/or depression, polyuria/polydipsia and vomiting. These signs are typical of AKI in dogs. ²⁵ , ²⁶ With the exception of polyuria/polydipsia, which was only reported in 55% of cases, the common signs are nonspecific and potentially attributable to another body system. For example, anorexia and vomiting could easily be misattributed to a gastrointestinal complication, thus delaying diagnosis. If clinical signs develop many days after the procedure, attributing them to perioperative renal injury may be even less likely. Considering the time course and potential for misattribution of clinical signs, it is possible that perioperative AKI is a substantially underappreciated cause of morbidity and mortality in dogs recovering from anaesthesia and surgery. Dogs should be monitored for potential signs of AKI, including nonspecific signs such as anorexia, lethargy or vomiting, for at least 4 days and ideally for 2 weeks postoperatively.

It is noteworthy that neither oliguria nor anuria, which are diagnostic features of AKI ¹⁵ were reported, whereas polyuria was common. The reasons for this are unknown, as many dogs had been discharged from hospital and then represented days later. It is possible that a brief (≤24 h) period of very low urine output went unnoticed. By the time clinical signs were observed, dogs may have progressed to a polyuric phase of AKI. ²⁵ Alternatively, low urine output may have been interpreted by attending veterinarians as physiologically appropriate in the context of an anorexic, dehydrated or vomiting dog. As neither polyuria nor oliguria were reported in almost half of the cases, lack of obvious changes in urine output should not preclude investigation for AKI in this context. ²⁵ , ²⁶

Female dogs appeared to be significantly overrepresented in the included reports, comprising 81% of the cases where sex was reported. In contrast, male sex is a known independent risk factor for perioperative AKI in adult humans even when common comorbidities such as obesity are accounted for. ⁴ However, major abdominal surgery within the peritoneal cavity is also a risk factor in humans ⁴ and female desexing procedures were the most common surgical procedures in the included cases (42%). This may have inherently confounded sex and procedure, as desexing surgery is one of the most common surgical procedures performed in female dogs. ²⁷

Large breed dogs with BW over 20 kg were identified as being at increased risk of perioperative AKI by Rogers‐Smith et al. ¹³ Results of this review provide further support to this finding, as the same trend is observed in the other 8 included reports. Furthermore, only one dog weighed less than 17 kg. It is plausible that surgery times may be longer in larger dogs thus increasing anaesthetic time and risk ²⁸ or that the larger volumes of fluid therapy required to correct dehydration or blood loss may have delayed effective therapy for hypovolaemia. It has also been speculated that larger dogs may require lower doses of drugs due to allometric scaling, placing them at increased risk of overdose. ¹³ Labradors and Golden Retrievers were the only breeds represented more than once; together, these two related breeds accounted for 23% of the cases. These breeds comprised 21% of hospital acquired AKI cases in one study of ICU patients. ²⁹ Most of these cases had undergone anaesthesia including all of the Labradors and Golden retrievers. In contrast, sporting breeds, which would include Labradors and Golden Retrievers, were at lower risk of AKI in an earlier case‐control study. ²⁶ It is not possible to determine from the current review if larger size per se was the risk factor or if these popular large breeds ³⁰ are at increased risk, thus increasing the proportion of larger dogs.

Perioperative AKI is a multifactorial complication ⁴ , ⁶ ; unsurprisingly, many potential causes of AKI were implicated in the included reports. Reduced renal perfusion due to combinations of dehydration, hypovolaemia and cardiovascular depression from anaesthetic drugs were suggested by all authors as a potential cause. Severe hypovolaemia and hypotension during anaesthesia have been shown to cause renal injury in dogs. ³¹ Blood pressure monitoring and intraoperative IV fluids were used in the majority of cases, and most authors did not report any hypotensive events. However, no definitions of hypotension were provided, complicating interpretation of these claims. If a time‐based definition were used, such as mean arterial pressure less than 60 mmHg for over 10 minutes, ³² short periods of severe hypotension may not have been reported. Likewise, in the cases series where two hypotensive events (descried as mean arterial pressure < 60 mmHg) were reported ¹³ the severity of these events is not recorded. Although IV fluids will not prevent hypotension during anaesthesia, ³³ both excessive and inadequate IV fluid therapy will contribute to AKI. ¹⁶ Consequently, it is difficult to exclude hypoperfusion as at least a contributing cause to these AKI cases. Nephrotoxins, including antibiotics nafcillin and trimethoprim sulphadiazine ¹⁴ , ²⁰ ; and the anaesthetic methoxyflurane ¹⁸ were implicated in several reports. Although most of these drugs have been superseded, other nephrotoxic substances, including sulphonamides or tetracycline antibiotics and iodinated contrast agents, remain in perioperative use. ⁵ , ¹⁶ Nonsteroidal Anti‐inflammatory Drugs were the most commonly used potential nephrotoxin. By inhibiting renal production of prostaglandins, NSAIDS inhibit a key protective mechanism for the kidneys when perfusion is impaired. ³⁴ , ³⁵ Although NSAIDs alone may not inherently cause renal injury, they can act synergistically with other insults to cause AKI, as was demonstrated in one study with the combination of flunixin and methoxyflurane. ¹⁸ Although the older NSAID flunixin was implicated in several reports, ¹⁸ , ²⁰ , ²¹ the newer cyclooxygenase‐2 (COX‐2) selective drug meloxicam was the most frequently used. ¹³ As prostaglandin production via COX‐2 is essential for maintaining renal perfusion, even highly COX‐2 selective NSAIDs, such as meloxicam, will still pose a risk to renal perfusion. ³⁴ , ³⁵

There are inherent limitations resulting from the design and scope of this review. Due to the focus on the development of overt clinical signs and mortality, only studies reporting dogs with relatively severe disease are included. Publications describing subclinical or biomarker positive AKI with normal creatinine were excluded. Findings from this review should not be extrapolated to these situations, and the prevalence of MAKE is likely to be higher in this population with severe disease compared with lower grade AKI. However, the purpose of the review was to provide information to assist with monitoring and thus early identification and treatment of dogs with severe disease.

Selection for publications that provide specific details about the time course of perioperative AKI will inherently select for smaller studies such as case series or case‐control studies where these details are reported. As a consequence, the total number of animals included in the review is small, limiting the power of the study; additionally, all included publications were retrospective. As case reports are more likely to relate unexpected outcomes, it is possible that the development of AKI in younger animals undergoing elective surgery (e.g. ovariohysterectomy) is overrepresented and results may differ in higher risk procedures. Findings relating to risk factors such as BW and sex, or potential causes of perioperative AKI should be treated as hypothesis generating only.

As no restriction on publication date was imposed, publications were included from across more than three decades. Some of the drugs used in the earlier reports are now obsolete and anaesthetic practices including monitoring and supportive care have presumably improved over the decades. Hopefully these changes will have resulted in a decrease in anaesthetic complications ³ including AKI over time. This would result in a lower incidence of perioperative AKI and perhaps a lower mortality rate if critical care practice has similarly evolved. However, the findings relating to the time course of actual AKI cases may not have changed. Chronological examination of the included reports would support this suggestion (Table 1). The time course of onset of clinical signs, reported signs and also the proportion of female and larger dogs is remarkably consistent between publications across more than 35 years. This consistency increases confidence in these main findings and suggests that the manner in which severe AKI presents postoperatively is the same regardless of the inciting causes.

Some recommendations for clinical practice that can be drawn from the findings of this review. Firstly, dogs should be monitored for potential signs of AKI postoperatively for at least 4 days and ideally for 2 weeks. Secondly, AKI should be considered as a differential diagnosis for anorexia, vomiting, lethargy, depression or dehydration in any dog that has undergone anaesthesia and surgery within 14 days. Veterinarians should consider measuring serum creatinine and urea, along with urine specific gravity in dogs that develop these signs postoperatively even if they do not have specific clinical signs such as polyuria or oliguria.

Conclusion

In dogs with severe perioperative AKI, clinical signs are typically first seen between two and 4 days postoperatively but may occur as early as 1 day or as late as 14. The most common clinical signs include anorexia, lethargy or depression, polyuria/polydipsia and vomiting. By the time clinical signs become apparent, urine is normally isosthenuric with normal to elevated output. Knowledge of this time course and clinical signs may improve postoperative monitoring and recognition of perioperative AKI in dogs. Mortality and other major adverse kidney events appear common, but further study is needed to determine the incidence of these complications. It is possible that female dogs, or at least those undergoing desexing surgery, are at increased risk of AKI. Larger breed dogs, especially Labradors and Golden Retrievers, also appear to be at increased risk. The possibility that sex, size and/or breed are risk factors should be evaluated by larger appropriately controlled studies.

Conflicts of interest and sources of funding

The author has no conflicts of interest to declare. Researcher time and academic library access was funded by Charles Sturt University. Funding for the included sources of evidence was not declared.

Quinn, C.T. , Time course of clinical signs and mortality in dogs with severe perioperative acute kidney injury: A scoping review. Aust Vet J. 2025;103:443–449. 10.1111/avj.13454

Data availability statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.