//
© 2026 Zoe. All rights reserved.
Zoe is a legal information platform. Always consult the Federal Register of Legislation for authoritative text.
[2026] NSWCA 71
respondent. Appeal dismissed. The appellant was ordered to pay the respondent’s costs of the appeal. The appellant was also ordered to pay the costs of the fresh-evidence application, but those costs did not...
The Court of Appeal upheld the damages judgment arising from a dog attack. It accepted that the trial judge could prefer the respondent’s psychiatrist, rely on contemporaneous treatment records, and find that the attack caused PTSD and some loss of earning capacity. The trial judge used the wrong label for the source of the causation test, but reached the correct result using the applicable statutory test. The appeal failed, as did the appellant’s attempt to rely on public material about the respondent’s later business activity.
AI-generated legal information, not legal advice. Zoe can make mistakes. Check the cited source, and for advice about your situation consult a qualified Australian lawyer.
9 cited instruments linked from this judgment.
2,648 words · generated 08/10/2026
On 24 April 2020, the respondent, Maria Ioannidis, was walking her shih tzu, Lexi, in Mortdale when a large dog owned by the appellant, Melissa Carretero, attacked them. The attack’s circumstances were not disputed. The attacking dog badly injured Lexi, ultimately requiring amputation of her leg at the shoulder, and bit the respondent on her neck, chin, arm and legs. The Court described it as “a terrifying attack” [3]-[5].
The respondent brought a claim under s 25(1)(a) of the Companion Animals Act 1998 (NSW) (CAA). That provision makes a dog’s owner liable in damages for bodily injury to a person caused by the dog wounding or attacking that person. Liability was common ground [5], [54]. The contested questions concerned the extent and cause of the respondent’s injuries and the resulting damages.
The respondent had a history of anxiety and depression, but said that her symptoms and day-to-day functioning changed substantially after the attack. Her medical records included early treatment for anxiety and trauma-related symptoms, a psychologist’s assessment of acute stress disorder, and later records referring to PTSD-like symptoms and PTSD [35]-[45]. The psychiatrists retained by the parties disagreed. Dr Chow considered that the respondent had PTSD caused by the attack and that her ability to work was affected [47]. Dr Lee considered that exaggeration or feigning could not be excluded and did not accept the PTSD diagnosis [48]-[53].
The respondent operated a couture fashion business. It remained active and its recorded sales increased after 2019, but she said she could no longer work the hours she had previously worked. Her partner described assisting her with transport, appointments and business communications [29]-[33]. The primary judge found some impairment of earning capacity, while recognising that the respondent probably worked more and had greater capacity than she claimed [63].
The District Court awarded $101,788.63: $49,500 for non-economic loss, $2,288.63 for agreed past expenses, $15,000 as a future-expenses buffer and $35,000 as a buffer for past and future loss of earning capacity [6]. On appeal, Carretero challenged the psychiatric findings, causation, the earning-capacity finding and the damages assessment. She also sought to introduce public articles and social-media material about Ioannidis’s business activity after the trial. The Court dismissed that fresh-evidence motion and then dismissed the appeal [10], [96]-[97].
The Court addressed the appellate standard first. It rejected the idea that one standard necessarily governs every component of a damages award. Essential factual findings, including whether the respondent suffered a genuine condition and whether it affected her earning capacity, were reviewed for correctness, while applying the appellate principles in Fox v Percy and Lee v Lee [65]-[66]. Causation also attracted correctness review because it was an inquiry with one legally correct outcome [66].
The assessment of non-economic loss was different. Assigning an injury a proportion of a “most extreme case” requires an evaluative judgment rather than the discovery of a single precise answer. The Court held that the House v The King standard was appropriate for that assessment [67]-[68]. For loss of earning capacity, the Court noted authority asking whether the award fell within the range available or a reasonable range, reflecting the imprecise nature of the exercise [69]. It added that the outcome would not change even if correctness were the test for that award [69].
On the psychiatric evidence, the Court agreed that Dr Chow’s opinion could be preferred. The contemporaneous medical records were important: they recorded anxiety and trauma-related symptoms soon after the attack, referrals for psychological care and later psychiatric treatment [35]-[45], [72]. The primary judge had also observed the respondent giving evidence and considered her partner’s supporting testimony [55], [72]. That advantage formed part of the evidence available to the trial judge.
The Court identified weaknesses in Dr Lee’s reasoning. His report omitted or did not address records that documented early psychological symptoms. In particular, he said there was no verification of PTSD symptoms until much later, despite records of treatment and an acute stress disorder assessment in 2020 [49]-[52]. Dr Chow, by contrast, set out the relevant history, his examination findings and his reasoning. The Court considered that the facts and process of reasoning were sufficiently disclosed [73]-[74]. It also explained that a psychiatrist may assess whether reported symptoms are genuine through the clinical assessment and history; symptom-validity testing is not the only possible method [74].
The causation issue involved a legal correction to the primary judge’s reasoning, but not a material error. The trial judge had said that common law, rather than s 5D of the Civil Liability Act 2002 (NSW), governed causation [57]. On appeal, Stern JA held that s 5D did not apply because s 25 liability does not depend on a breach of duty. The causation question instead arose under the CAA’s words “caused by”, interpreted by reference to that statute’s text and purpose [78]-[79]. The Court nevertheless considered common-law causation principles informative in this case and found that the primary judge had applied the correct substantive approach [79].
The proposed alternative causes included the respondent’s pre-existing mental-health history, COVID-19 impacts on her business, earlier life events and her father’s death [77]. The Court did not require the primary judge to investigate each possibility in the abstract. The alternatives had not been put to the respondent in cross-examination, and neither psychiatric expert attributed her PTSD to them [81]-[82]. The evidence instead linked the symptoms over time to the dog attack. The Court therefore rejected the complaint that the trial judge had relied merely on temporal association [80]-[83].
The earning-capacity finding survived for similar evidentiary reasons. The primary judge had not accepted the respondent’s account without qualification. He considered the financial records, business performance, gym attendance, medical notes and the partner’s evidence, and recognised that the respondent might have overstated the extent of her incapacity [59], [63]. Even so, the evidence supported some reduction in capacity. The Court held that the finding was neither “glaringly improbable” nor contrary to established facts [85]-[87]. It also rejected the proposition that the judge had to identify particular tasks the respondent could not perform, given the evidence of broader symptoms such as difficulty leaving home, anxiety, low motivation and depression [87].
The fresh-evidence application failed on two independent grounds. The material pre-dated trial and was publicly available, so it could have been found with reasonable diligence. The appellant’s own trial cross-examination had used online material, demonstrating the practical ability to conduct searches [17]-[18]. Further, the proposed material was unlikely to change the result. The respondent had acknowledged that she remained active in business and had explained her social-media use as part of promoting it. The material did not disprove her claim that she could not work full time [19]-[20].
The judgment’s principal clarification concerns appellate review of damages assessments. Before this decision, the authorities discussed in the judgment reflected different approaches to reviewing findings about the severity of non-economic loss. Stern JA identified a possible tension between Hall v State of New South Wales and White v Redding, as described in Stanberg v State of New South Wales [67]. The Court preferred the analysis in White: a proportion-of-a-most-extreme-case assessment is impressionistic, so appellate review should follow House v The King rather than require one uniquely correct numerical assessment [67]-[68].
The judgment also emphasises that the standard of review is not necessarily uniform across a personal-injury award. The Court applied correctness review to the relevant factual and causation findings, while treating the non-economic-loss assessment as a range-tolerant evaluative judgment. It discussed a reasonable-range approach to loss of earning capacity, while noting that the result would be the same under correctness review [65]-[69]. The decision therefore presents the task as classification of the particular finding under challenge, rather than applying one label to the entire damages award.
For dog-attack claims, the judgment clarifies that s 5D of the Civil Liability Act does not govern a claim under s 25(1)(a) of the CAA. The reason is that the CAA liability considered here does not depend on a breach of duty [78]. The statutory words “caused by” supply the relevant test, interpreted by reference to the text and purpose of the CAA [79]. The primary judge’s contrary description of the source of the test was an error, but it did not invalidate the result because the correct causation principles were applied in substance [79]-[80].
The judgment also explains the evidentiary context in which a trial judge may consider alternative causes of psychiatric injury. It does not establish that alternative causes are generally irrelevant. Rather, on these facts, the judge was not required to pursue possibilities that were not put to the witness and were not advanced by either expert as causes or contributors to the diagnosed PTSD [81]-[82]. The limits of that reasoning depend on the evidence and the way the case is presented at trial.
Finally, the decision illustrates application of the “special grounds” requirement for fresh evidence on appeal from a trial judgment. It restates the general conditions identified in Dibb: reasonable diligence, likely important influence on the result, and apparent credibility [11]-[12]. The application failed on diligence and likely influence, each being independently fatal [17]-[20].
At [68], Stern JA said that it would be “wholly artificial” to treat the assessment of non-economic loss as having one uniquely correct answer. In plain English, appellate judges recognise that evaluating pain, suffering and other non-economic effects involves judgment and approximation. An appeal does not succeed merely because another judge might have chosen a different percentage, provided the trial judge’s assessment does not meet the House v The King threshold for intervention.
At [79], the Court said that causation under s 25(1)(a) should be determined by reference to the CAA’s text and purpose. This means the source of the test is the statute creating the dog owner’s liability, not an automatic application of negligence legislation. In this case, common-law principles helped inform the statutory inquiry, but the Court distinguished that approach from treating s 5D as controlling.
At [82], the Court held that the primary judge did not have to search the evidence for every event that might theoretically cause anxiety. The important limits were that the proposed events were not put to the respondent and neither expert said they caused or contributed to the PTSD. The passage is a fact-specific explanation of when an asserted alternative cause may be too speculative to require separate analysis.
At [85], the Court applied appellate restraint to the trial judge’s finding of impaired earning capacity. The phrase “neither glaringly improbable nor contrary to the established facts” conveys that an appeal court does not simply substitute its own view of contested evidence. The trial judge had heard the witnesses and considered the surrounding records; the finding remained open on that evidence.
At [92], the Court described the $35,000 earning-capacity buffer as conservative and said it adequately reflected identified uncertainties. That conclusion depended on the respondent’s potential working life, the period affected by pandemic restrictions, the business’s capacity to generate income and the evidence of impaired capacity. It is not a general tariff for similar cases.
The decision is most directly relevant where a personal-injury damages appeal challenges different parts of the award, rather than treating the assessment as one indivisible exercise. A party disputing factual findings about injury or causation should expect those questions to be considered under the correctness standard, informed by appellate restraint principles. A challenge to the proportion assigned to non-economic loss engages the House v The King approach identified at [68]. A challenge to an earning-capacity buffer raises the reasonable-range authorities discussed at [69], although the judgment cautions that classification may not affect the result in every case.
The CAA causation reasoning is relevant where a claimant seeks damages for bodily injury caused by a dog wounding or attacking them under s 25(1)(a). The decision indicates that s 5D of the Civil Liability Act is not the governing provision for that statutory claim [78]-[79]. It will be especially relevant where the parties dispute whether a psychiatric condition was caused by the attack or by earlier or later events.
The alternative-cause discussion is most relevant where a party asserts that other stressors contributed to psychiatric symptoms. This judgment involved alternatives that were not put to the respondent and were not identified by the experts as causes of her PTSD [81]-[82]. It should not be read as excusing analysis of alternative causes where the evidence raises them directly, the witness is questioned about them, or expert evidence supports them.
The fresh-evidence reasoning is relevant to applications to introduce pre-trial material after a trial judgment. Public availability, the applicant’s ability to search for similar material before trial, and whether the material could materially affect the outcome were central here [17]-[20]. The judgment does not say that publicly available material can never be fresh evidence; it explains why this particular material failed the statutory conditions.
The supplied judgment contains no account of later decisions applying, distinguishing or otherwise treating Carretero v Ioannidis (No 3). It is therefore not possible, on the material provided, to describe its subsequent treatment or to claim that later courts have adopted its reasoning.
The judgment itself relies on prior authority rather than later cases. Its appellate-review analysis draws on Harper, Diamond, House, White, Hall and the High Court’s discussion of correctness and range-tolerant outcomes [65]-[69]. Its explanation of CAA causation relies on Simon v Condran and Wallace v Kam [78]. The fresh-evidence principles are summarised from Dibb [12]. Those citations show the authorities the Court considered relevant when deciding this appeal; they do not establish how later courts have treated this judgment.
The judgment identifies, but does not finally resolve for every type of damages finding, how the line between correctness review and House v The King review should be drawn. Stern JA observed that the line between a criterion demanding a unique answer and one tolerating a range is “not bright” [68]. The judgment’s approach is to identify the character of the particular finding, but it does not provide a single formula for classifying every economic or non-economic assessment.
The Court also did not have to decide whether the same approach to statutory causation under s 25(1)(a) would apply to materially different injury mechanisms or evidentiary disputes. It concluded that common-law principles informed the statutory test in this case and that the primary judge’s error was immaterial [79]. The decision does not comprehensively define the operation of “caused by” for all claims under the CAA.
The treatment of competing causes is similarly confined to the circumstances before the Court. The alternatives were not put to the respondent, and neither expert attributed the PTSD to them [81]-[82]. The decision leaves open how a trial judge should analyse multiple plausible contributors where those matters are squarely raised in evidence or supported by expert opinion.
The earning-capacity discussion does not establish when a claimant’s business revenue will prove or disprove impairment. Here, the Court considered the financial records alongside evidence that the respondent worked fewer hours, had support from her partner and experienced psychological symptoms affecting her activities [29]-[33], [91]-[92]. The result depended on that combination, not on revenue figures alone.
The judgment also does not decide that online material is inherently insufficient to establish a person’s work capacity. It found that the particular articles and videos did not materially contradict the respondent’s trial evidence, which already acknowledged continued business activity [19]-[20]. Whether comparable material would matter in another case would depend on its content, timing, context and relationship to the disputed capacity.
This is general legal information, not legal advice. Zoe can make mistakes. Before acting on this, relying on it, or using any draft, have a qualified Australian lawyer check it.
Source: NSW Caselaw (www.caselaw.nsw.gov.au). Converted to plain text by ZoeLaw. Not the official version.
Medium Neutral Citation: Carretero v Ioannidis (No 3) [2026] NSWCA 71
Hearing dates: 24 March 2026
Date of orders: 28 April 2026
Decision date: 28 April 2026
Before: Bell CJ at [1];
Kirk JA at [2];
Stern JA at [3]
Decision: (1) Appeal dismissed.
(2) Appellant to pay the respondent’s costs of the appeal.
Catchwords: ANIMALS — companion animals — dogs — where respondent and her dog were attacked by appellant’s dog — where respondent sued under s 25 Companion Animals Act 1998 (NSW) — where circumstances of attack not in dispute — where primary judge gave judgment for respondent — where primary judge found respondent suffered both physical and psychiatric injury — whether primary judge erred in preferring opinion of one psychiatric expert over another — whether primary judge erred in approach to causation — whether evidence before primary judge did not support his Honour’s finding that respondent had suffered impairment of earning capacity — whether primary judge erred in assessment of damages — appeal dismissed
Legislation Cited: Civil Liability Act 2002 (NSW), s 5D
Companion Animals Act 1998 (NSW), s 25
Evidence Act 1995 (NSW), s 79
Supreme Court Act 1970 (NSW), s 75A
Uniform Civil Procedure Rules 2005 (NSW), r 50.20
Cases Cited: Australian Securities & Investment Commission v Rich [2005] NSWCA 152; 218 ALR 764
Brear v James Hardie & Coy Pty Ltd (2000) 50 NSWLR 388; [2000] NSWCA 352
Chen by her tutor Huang v Kmart Australia Ltd [2023] NSWCA 96
Comcare v Martin (2016) 258 CLR 467; [2016] HCA 43
Diamond v Simpson (No 1) [2003] NSWCA 67
Dibb v Transport for New South Wales [2024] NSWCA 157
Fisher v Nonconformist Pty Ltd (2024) 114 NSWLR 1; [2024] NSWCA 32
Fox v Percy (2003) 214 CLR 118; [2003] HCA 22
GLJ v Trustees of the Roman Catholic Church for the Diocese of Lismore (2023) 280 CLR 442; [2023] HCA 32
Hall v State of New South Wales [2014] NSWCA 154
Harper v Bangalow Motors Pty Ltd [1990] NSWCA 85
House v The King (1936) 55 CLR 499; [1936] HCA 40
Ioannidis v Carretero [2025] NSWDC 258
J and E Vella Pty Ltd v Hobson [2023] NSWCA 234
Langdon v Carnival PLC [2024] NSWCA 168
Lee v Lee (2019) 266 CLR 129; [2019] HCA 28
Loxton v State of New South Wales [2002] NSWCA 194
Makita (Australia) Pty Ltd v Sprowles (2001) 52 NSWLR 705; [2001] NSWCA 305
Miller v Jennings (1954) 92 CLR 190; [1954] HCA 65
Minister for Immigration v SZVFW (2018) 264 CLR 541; [2018] HCA 30
Moore (a pseudonym) v The King [2024] HCA 30; 98 ALJR 1119
Motor Accident Commission v Raccanello [2025] SASCA 146
Pollard v Baulderstone Hornibrook Engineering Pty Ltd [2008] NSWCA 99
Ramsay v Watson (1961) 108 CLR 642; [1961] HCA 65
Seltsam Pty Ltd v Ghaleb [2005] NSWCA 208
Simon v Condran (2013) 85 NSWLR 768; [2013] NSWCA 388
Stanberg v State of New South Wales [2025] NSWCA 127
Transport Accident Commission v Pastuch [2025] SASCA 120
Wallace v Kam (2013) 250 CLR 375; [2013] HCA 19
Warren v Coombes (1979) 142 CLR 531; [1979] HCA 9
White v Redding (2019) 99 NSWLR 605; [2019] NSWCA 152
Category: Principal judgment
Parties: Melissa Carretero (Appellant / Applicant on Motion filed 26 November 2025)
Maria Ioannidis (Respondent / Respondent on Motion filed 26 November 2025)
Representation: Counsel:
In person (Appellant / Applicant on Motion filed 26 November 2025)
LD Robison (Respondent / Respondent on Motion filed 26 November 2025)
Solicitors:
Shine Lawyers (Respondent / Respondent on Motion filed 26 November 2025)
File Number(s): 2025/319946
Publication restriction: Nil
[This headnote is not to be read as part of the judgment]
On 24 April 2020 the respondent and her dog were attacked by the appellant’s dog. The circumstances of the attack were not in dispute. The respondent’s dog’s paw was ripped off at the joint and the respondent was bitten on the neck, chin, arm and legs. It was common ground that s 25(1)(a) of the Companion Animals Act 1998 (NSW) (CAA), which provides that the owner of a dog is liable in damages in respect of bodily injury to a person caused by the dog wounding or attacking that person, was engaged. The primary judge gave judgment for the respondent in the sum of $101,788.63. His Honour found that the respondent suffered both physical injury as well as psychiatric injury in the form of post-traumatic stress disorder (PTSD) as a result of the attack.
The key issues before this Court were:
(i) whether the primary judge erred in preferring the opinion of the respondent’s psychiatric expert, Dr Chow, over that of the appellant’s, Dr Lee;
(ii) whether the primary judge erred in his approach to causation;
(iii) whether the evidence before the primary judge did not support his Honour’s finding that the respondent had suffered impairment of earning capacity; and
(iv) whether the primary judge erred in his assessment of damages.
The Court (Stern JA, Bell CJ and Kirk JA agreeing) held, dismissing the appeal:
As to the applicable standard of appellate review
(1) The same standard of appellate review will not automatically be applied to all elements, economic and non-economic, of a claim for damages: [65].
Harper v Bangalow Motors Pty Ltd [1990] NSWCA 85, Diamond v Simpson (No 1) [2003] NSWCA 67, applied.
(2) The House v The King standard of appellate review is applicable as regards an assessment of non-economic loss as it would be wholly artificial to describe the exercise as one in which there is only one legally correct answer or a unique outcome: [68].
Minister for Immigration v SZVFW (2018) 264 CLR 541; [2018] HCA 30, GLJ v Trustees of the Roman Catholic Church for the Diocese of Lismore (2023) 280 CLR 442; [2023] HCA 32, Moore (a pseudonym) v The King [2024] HCA 30; (2024) 98 ALJR 1119, applied.
(3) There is a line of authority that indicates that the question for the appellate court is whether the amount awarded was within the range available to the trial judge or within “a reasonable range”, and this reflects the necessarily impressionistic nature of the exercise: [69].
Loxton v State of New South Wales [2002] NSWCA 194, Pollard v Baulderstone Hornibrook Engineering Pty Ltd [2008] NSWCA 99, Brear v James Hardie & Coy Pty Ltd (2000) 50 NSWLR 388, [2000] NSWCA 352, applied.
As to issue (i)
(4) The primary judge was correct to prefer Dr Chow’s opinion over that of Dr Lee. Both the facts relied upon by Dr Chow and his process of reasoning were sufficiently disclosed in his report: [70]-[75].
As to issue (ii)
(5) Section 5D of the Civil Liability Act 2002 (NSW) did not apply in the context of the claim under s 25 of the CAA. The statutory test for causation under s 25(1)(a) of the CAA applied. The primary judge erred in holding that the common law applied but nonetheless applied the correct principles to determine causation: [78]-[79].
Simon v Condran (2013) 85 NSWLR 768; [2013] NSWCA 388, Wallace v Kam (2013) 250 CLR 375; [2013] HCA 19, applied.
(6) The primary judge was not required to scrutinise the material before him to ascertain whether there might have been competing causes that may have led to anxiety or a stress reaction, in circumstances where these were not put to the respondent and neither of the experts suggested that any of the posited competing causes had caused or contributed to her PTSD: [66], [80]-[83].
As to issue (iii)
(7) His Honour’s finding that the respondent’s earning capacity was impaired was neither glaringly improbable nor contrary to the established facts: [66], [84]-[85].
Fox v Percy (2003) 214 CLR 118; [2003] HCA 22, Lee v Lee (2019) 266 CLR 129; [2019] HCA 28, J and E Vella Pty Ltd v Hobson [2023] NSWCA 234, applied.
As to issue (iv)
(8) The buffer for impairment of earning capacity awarded more than adequately gave effect to the uncertainties reflected in the evidence and the primary judge’s assessment of non-economic loss and allowance for future treatment expenses were not erroneous: [69], [89]-[93].
BELL CJ: I agree with Stern JA.
KIRK JA: I agree with Stern JA.
STERN JA**:** On 24 April 2020, at around 6 pm, the respondent and her shih tzu dog Lexi were attacked while walking in Mortdale by a large greyhound staffordshire bull terrier cross dog owned by the appellant. In what was, on any view, a terrifying attack, Lexi’s paw was ripped off at the joint (requiring amputation of the leg at the shoulder) and the respondent was bitten on the neck, chin, arm and legs. The appellant’s dog was put down. The respondent sued the appellant under s 25 of the Companion Animals Act 1998 (NSW) (CAA).
The circumstances of the attack are not in dispute. As the primary judge explained them:
[T]he [appellant’s] dog, a large greyhound staffordshire bull terrier cross ran down the street unrestrained and viciously attacked the [respondent’s] dog. The [respondent] described graphically how the [appellant’s] dog grabbed Lexi and was throwing her from side to side as the [respondent] tried desperately to pick her up. When she finally was able to get Lexi, the [appellant’s] dog continued to attack. In the [respondent’s] words:
“… he kept running back and launching at me and he just kept attacking, and he just didn’t stop. He didn’t stop, and I was falling over and trying to get on my feet.”
The [respondent] described how whilst she was holding Lexi the [appellant’s] dog ripped Lexi’s whole paw off at the joint. She recounted how the dog bit the [respondent] across the neck, her face around the chin, her arm and her legs. The [respondent] described fingernails being completely ripped off in the chaos of the attack as she was falling over and tripping on the lead, all the time trying not to let go of Lexi.
(1) The owner of a dog is liable in damages in respect of—
(a) bodily injury to a person caused by the dog wounding or attacking that person, …
On 15 July 2025 the primary judge gave judgment for the respondent in the sum of $101,788.63 and ordered that the appellant pay the respondent’s costs. The damages comprised $49,500 for non-economic loss, $2,288.63 for past out of pocket expenses, $15,000 as a buffer for future out of pocket expenses and $35,000 as a buffer for past and future loss of earnings: Ioannidis v Carretero [2025] NSWDC 258. The primary judge found that the respondent suffered both physical and psychiatric injury in the form of post-traumatic stress disorder (PTSD) as a result of the attack in the context of pre-existing anxiety and depression.
The appellant now appeals under s 75A of the Supreme Court Act 1970 (NSW) against the primary judge’s assessment of damages. She challenges the primary judge’s findings as to the causation and extent of the respondent’s psychiatric injury and loss of earning capacity and his Honour’s assessment of non-economic loss. She contends, in essence, that:
the primary judge did not properly consider the report of her psychiatric expert, Dr Leonard Lee (g****round o****ne);
the primary judge failed properly to evaluate potential contributing causes of the respondent’s symptoms other than the attack, which the appellant says was required of his Honour by s 5D of the Civil Liability Act 2002 (NSW) (CLA) (g****round t****wo);
the primary judge erred in finding that the respondent had little or no earning capacity (g****round t****hree);
the primary judge’s assessments of both economic and non-economic loss were manifestly excessive (g****round f****our);
key factual findings made by the primary judge were inconsistent with the documentary and expert evidence (g****round f****ive);
fresh evidence demonstrates that the respondent had misled the primary judge regarding her capacity to work and the severity of her psychological condition (ground six); and
the primary judge failed adequately to consider other causes of the respondent’s symptoms (g****round s****even).
By way of context, the appellant accepts that the respondent suffered some level of emotional reaction following the attack, which the appellant says was a confronting and upsetting event. It should be noted, however, that this description fails to reflect that, as the primary judge found, this was a violent and terrifying attack.
By notice of motion filed 26 November 2025, the appellant sought leave to adduce fresh evidence under s 75A(7) of the Supreme Court Act and Uniform Civil Procedure Rules 2005 (NSW), r 50.20. That application was considered, and dismissed, at the commencement of the hearing of the appeal and it follows that ground six falls away. Short-form reasons for that dismissal were given orally and it was noted that more fulsome reasons would be given in our judgment. Those reasons are set out below.
Further evidence may be received in appeals to this Court: Supreme Court Act, s 75A(7). However, where, as here, an appeal is from a judgment after trial, the Court shall not receive such evidence except on special grounds: s 75A(8). That prohibition does not apply to evidence concerning matters occurring after the trial: s 75A(9), but it was common ground that the material sought to adduced by the appellant concerned matters occurring prior to the trial.
The proper approach to these provisions was summarised in Dibb v Transport for New South Wales [2024] NSWCA 157 at [33]-[34]:
The expression “special grounds” has been adopted to “describe the circumstances which in a particular case will justify resolving the tension between the demands of justice on the one hand and the public interest in bringing suits to a final end on the other, in favour of allowing further evidence to be led”: Tjiong v Tjiong [2012] NSWCA 201 at [165] (Meagher JA, Whealy and Barrett JJA agreeing) (“Tjiong”) citing Phoenix Commercial Enterprises Pty Ltd v City of Canada Bay Council [2010] NSWCA 64 (Spiegelman CJ at [10], [22], [27], Campbell JA at [136], Handley AJA at [250]) (“Phoenix”).
In general, three conditions must be satisfied before further evidence will be received on “special grounds” under s 75A(8) of the Supreme Court Act:
(1) It must be shown that the evidence could not have been obtained with reasonable diligence for use at the trial: see eg Comlin Holdings Pty Ltd v Metlej Developments Pty Ltd (2019) 99 NSWLR 447; [2019] NSWCA 73 at [9]-[14] ("Comlin”);
(2) The evidence must be such that, if given, it would probably have an important influence on the result of the case, though it need not be decisive; and
(3) The evidence must be such as is presumably believed or, in other words, it must be apparently credible though it need not be incontrovertible.
The material that the appellant sought to adduce as fresh evidence comprised:
an article dated 13 June 2024, posted on a website www.ragtrader.com.au, reporting that an Australian swimwear label, Infamous Swim, had a group of new investors including the respondent and her partner, Matthew Child;
an article dated 21 May 2024, posted on a website www.smartcompany.com.au, describing the respondent and Mr Child as two members of a group of external investors coming on board to back Infamous Swim;
an article dated 12 June 2024, posted on a website www.insideretail.com.au, identifying new directors of Infamous Swim, including the respondent, Mr Child, and three others, and that the CEO of Infamous Swim would continue in her role;
an undated article, on a website www.revivalrunway.com, which the appellant says was published in 2024, reporting that Infamous Swim was one of Australia’s top swimwear and resortwear brands and that it had been brought out by a local group of self-funded entrepreneurs;
an article dated 3 October 2023, posted on a website www.smartcompany.com.au, reporting that Infamous Swim had generated $16.5 million in revenue in the past 12 months and was preparing to open up an offer to investors through a crowdfunding platform;
a video, posted on the @infamous_swim Instagram account on 2 June 2024, sharing the news that the respondent was a brand new partner of Infamous Swim, and depicting the respondent interacting with the Chief Executive Officer (CEO) of Infamous Swim and one other person and offering her opinion on swimwear in what appears to be a room with an Infamous Swim logo and a number of outfits in the background;
a video, posted on the @infamous_swim Instagram account on 20 August 2024, depicting the respondent interacting with models trying on swimwear in what looks like a dressing room but might be a fabricated background;
a video, posted on the @infamous_swim Instagram account on 31 October 2024, with the caption “[w]hen @maryioannidiscouture is in town”, depicting the respondent interacting animatedly with one other person in what looks to be the same room shown in the video at [13(6)] above;
an article dated 31 May 2023, posted on the website www.infamousswim.com.au, notifying of a Melbourne warehouse sale at 5/533 Frankston-Dandenong Road, South Dandenong, Victoria; and
an undated posting, on the website , notifying that a unit at 5/533 Frankston-Dandenong Road, Dandenong South, Victoria, one photograph on which looks visually similar to the background in the videos at [13(6)] and [13(8)] above, is available for lease.
In those circumstances I was not in a position to investigate or analyse the material in a way that would have revealed its significance. I relied on my partner and my legal representatives to manage communications during that period. Following the delivery of judgment I reviewed publicly available material relevant to the judgment and accidentally found material that I identified information that was inconsistent with what had been said at trial.
In support of this contention she relied upon:
a letter dated 4 November 2025 addressed to the “Supreme Court of NSW, Court of Appeal” from Michael Short, her solicitor-advocate in the first instance proceedings, indicating that on 28 and 30 June 2024 the appellant had sent emails to his office advising she would not take any calls “alone” moving forward to avoid getting upset, asking him to contact someone called Lachlan Williams for all required communication, and stating that the appellant had said this request was to protect both her mental health and the health and safety of her unborn child;
a letter dated 5 September 2024 from Dr Amy Edler at TasWomen stating that the appellant was pregnant with a due date of 5 January 2025 and asking for the trial to be deferred until July 2025 “to accommodate post-partum medical issues”; and
a letter dated 13 October 2025 from Kristina Suatoni, psychologist at Family Services Australia, identifying that she had been providing psychological treatment to the appellant since March 2024, that the appellant “reported experiencing significant emotional distress associated with ongoing court proceedings and the uncertainty surrounding their outcome” and that the appellant’s pregnancy, preparing for the sale of her home and “possible bankruptcy as a result of these legal matters” compounded this stress.
There were two principal reasons why the application to adduce fresh evidence was dismissed, each of which was, individually, fatal to the success of the application. First, all of the material sought to be adduced pre-dated the hearing before the primary judge and could readily have been obtained with reasonable diligence for use at that hearing. It was all publicly available material, having been identified by the appellant from searches which she explained she undertook following delivery of the primary judgment. The evidence fell well short of establishing that the appellant could not, herself, have carried out those searches prior to the trial which took place on 24 and 25 March 2025, being two and a half months after the appellant’s baby was due. In any event, the appellant was represented by solicitors at trial, and she advanced no reason why her solicitors could not have undertaken searches sufficient to locate the material.
It was, moreover, apparent that either the appellant, her solicitors, or someone else, did undertake Google and Instagram searches for the purpose of the respondent’s cross-examination at trial. Mr Short cross-examined the respondent by reference to posts on Instagram, media articles about her business and what was described as a “social media publication” posted on the website www.stuff.co.nz. Mr Short put to the respondent during cross-examination that he could show her “dozens of internet posts” through 2022 to the current time. This demonstrated that at that time the appellant, whether through her legal team or otherwise, was well able to undertake searches such as those which yielded the material now sought to be tendered as fresh evidence.
The respondent’s physical injuries involved pain associated with the attack and feeling sore for a few months afterwards. She acknowledged that she had no ongoing physical problems and only very minor scarring. She said, however, that she was traumatised by the incident and continued to struggle emotionally. She also described difficulty walking outside with her new dogs (Lexi having died in September 2024) and that, after the incident, she found it “really hard to leave the house” and “didn’t want to have to deal with any people”. She said that before the incident she would go out for dinner most nights but no longer had a social life. She also said that she used to play soccer but that “mentally” she could not do that any more. She still went to the gym in the mornings, and saw gym people there, but apart from that did not see friends. She also said that she used to be very house proud and did the shopping, cleaning, laundry, and most of the cooking but that, since the attack, her partner looked after all of these tasks. She said that she would still attend “weddings with family” but found that these exacerbated her anxiety.
The respondent gave evidence that she had had some mental health issues prior to the attack, including a history of anxiety, and also had migraines (which affected her sleep). She was under the care of her local doctor for this. She had consulted her doctor on account of a variety of matters including anxiety on 9 March 2020, approximately six weeks before the attack. She said, however, that after the attack her anxiety was “completely different”, explaining that prior to the attack she would “live [her] life”, leaving the house, meeting friends, going to dinner, and going to work every day.
Mr Child largely confirmed this evidence, explaining that his relationship with the respondent “changed quite dramatically on [24 April 2020]”. He said that after the attack she “doesn't want to leave the house, go anywhere, do anything, whereas she used to have pretty much normal time doing that sort of thing”. He agreed that the respondent had had issues with anxiety prior to the attack but that he “didn’t see it as really too bad” and she had “lived her normal life”. By contrast, he said that “now” she locked herself in the house and it was very hard to persuade her to leave.
The respondent ran her own business designing couture high-end bridal and evening wear (and that business continued at the time of trial). She set that business up in 1995 and she was the sole shareholder and sole director. She employed one contractor. She explained that she dressed a lot of celebrities. At the time of the attack, restrictions associated with the COVID-19 pandemic had impacted her business as weddings were shut down. She said prior to that she would go into work most days from Tuesday to Saturday and she had thought she would work until she was 70 or 80 years old. She said that at the time of trial, she tried to get into work maybe eight hours a week and did not know if she would still manage to work until the age at which she had previously anticipated retiring, as she struggled to go in for two half days a week. She said in her evidence that her business was “definitely active” but had suffered a downturn.
That's fair enough, but I would be making more if I could go into work five days or six days a week[.]
A number of posts from the respondent’s Instagram profile were put to her in cross-examination. She said that she frequently posted to social media for her business. She said that she had to try to get media exposure, explaining, “How will I survive if I don’t go in there?”.
It was put to the respondent that her evidence as to her reduced capacity to work was a fabrication. She denied that.
The respondent’s clinical records from three different medical centres were tendered at trial. Given the grounds of appeal, these are of some significance on the appeal.
On 25 April 2020, the day following the attack, the respondent sought treatment from Dr Fauzia Rehan at MedFirst Medical Centre, Hurstville. This was recorded as being the respondent’s first visit to that practice. Dr Rehan recorded a history of the attack and described the reason for visit as “[d]epression/Anxiety” and “[d]og scratch”. The respondent was given prescriptions for Keflex (an antibiotic) and temazepam (a benzodiazepine).
On 30 April 2020 the respondent saw Dr Mina Habashi at Hurstville Highpoint Medical Centre where she had been a patient since at least 2018. Dr Habashi recorded that the respondent had gone to see another general practitioner “on Saturday as this clinic was closed” (25 April 2020 having been a Saturday). He recorded that she had “witnessed a very traumatic [scene]”, described what she saw and then noted that, since then she had not been able to sleep. Dr Habashi advised the respondent to see a psychologist and completed a Mental Health Care Plan. The Mental Health Care Plan noted a history of anxiety, chronic migraine, cluster headache and SVT (which I presume refers to supraventricular tachycardia), that the presenting issue was “[s]ituational crisis witnessed her dog being viciously attacked”, and recorded a diagnosis of situational crisis and anxiety.
On 7 May 2020 the respondent saw Wai Lun Fu, psychologist. In a letter of that date to Dr Habashi, Mr Fu said:
Based on the symptoms she has reported, I believe she meets the criteria for Acute Stress Disorder directly as a result of the traumatic experience of being bitten and watching another dog mangle her own dog. She disclosed having experiences akin to flashbacks as well as nightmares and intrusive thoughts.
The notes of a telephone consultation with Dr Habashi on 9 July 2020 record that the respondent had been seeing a psychologist “to good effect” and a diagnosis of migraine and anxiety. On 24 June 2021 the respondent attended a consultation with Dr Wang. The notes of that consultation include reference to a “[history] of anxiety/panic attack” and a discussion about anxiety. The reasons for the visit were anxiety and to get a COVID-19 vaccination.
The respondent saw Dr Siyu Liu on 27 October 2021. Dr Liu’s notes record that the respondent had chronic anxiety which had been worse in the past two years since a dog attack in early 2020 and that the respondent also seemed to have “PTSD-like symptoms”. Dr Liu referred the respondent to Dr Wendy Orlay, psychiatrist and prescribed a trial of 50mg daily of Zoloft (an antidepressant). In the referral of 27 October 2021 to Dr Orlay, Dr Liu wrote:
Thank you for seeing Maria Ioannidis for a routine review and ongoing management. I have seen her a handful of times over the past 2 years. She has a history of chronic anxiety from her teens. Her dog suffered a dog attack in early 2020 and Maria is suffering from PTSD like symptoms which have been getting worse recently and severely affecting her quality of life. I have given her a script for Zoloft 50mg to start.
On 3 March 2022 the respondent saw Dr Monica Therese Tay complaining of chronic headaches, hyperhidrosis (excessive sweating) and poor sleep which were both said to be “likely in context of anxiety”, and anxiety which she reported as having happened after the attack. She had not taken up the previous referral to Dr Orlay and reported that she did not feel that she “clicked” with the psychologist who she had been seeing. There was a discussion about her returning for a Mental Health Care Plan.
On 14 March 2022 the respondent saw Dr Mark Wong who recorded a history, including reference to the attack and “[t]rauma”, and that the respondent “won[’]t leave house”, “won[’]t travel”. He wrote a letter addressed “[t]o whom it may concern”, in which he described the respondent as suffering from PTSD and said that Lexi should be designated the respondent’s “companion animal”. At a further consultation with Dr Wong on 30 March 2022 the respondent is recorded as saying that her anxiety/depression was “already a lot better with [P]ristiq”.
The notes of a further consultation with Dr Wong on 7 September 2022 record that the respondent’s anxiety was “very bad” and that she was struggling to get out of the house but “[o]k getting to work, with difficulty”. Even walking down the street she was “constantly … on edge”. The notes refer to the attack and that “even seeing a dog freaks [the respondent] out”. Her dose of Pristiq was increased and Dr Wong wrote a letter of referral to a psychologist, giving a history of “PTSD with intense flashbacks, since her dog was grievously mauled 4f20” and describing “[w]orsening symptoms with agoraphobia developing. Hypervigilance, poor sleep – poor concentration, struggling at her own work”. He also said that the respondent reported that she was “able to still go to her own workplace but it can be hard with her ptsd”. Under the heading “Past History”, there is no mention of any psychiatric illness until April 2020, where “PTSD – Dog attack” is noted.
The respondent had another consultation with Dr Wong on 7 October 2022. The notes record the respondent was “[c]oping” and was still able to go to her own workplace, but that “it can be hard with her ptsd”.
The respondent attended Dr Wong on 6 February 2023. Dr Wong noted that the respondent still “gets triggered ptsd” was “becoming obsessed with Lexi 12 year old”, “cant go overseas” and “still manages to get to work, struggles to get there”. The notes of a further consultation with Dr Wong on 24 August 2023 include reference to “nausea”, “tiredness”, and “shakiness” and said that she “gets depressed when home [alone]” and “goes to gym, work without trouble”. The records of a further consultation with Dr Wong on 13 September 2023 include, under the heading, “Reason for visit”, “PTSD” and also that the respondent “hates meeting new people”. Dr Wong wrote a letter “re. Names of psychologists” and recorded “for mha again once appt made”. On that day Dr Wong wrote a letter of referral to Dr Sim Gulati, consultant psychiatrist, asking him to please “consider all options including some of the newer treatments” and describing the respondent as suffering from “SEVERE PTSD”. He noted that the DASS-21 had indicated extremely severe D (depression), A (anxiety), and S (stress). Dr Gulati saw the respondent on 14 February 2023 and recorded that his impression was:
Maria is a 52yo female who presents with significant symptoms of PTSD including increased anxiety, irritability, nightmares, startle response and flashbacks. She has received some benefit from desvenlafaxine [Pristiq] which should be continued at this stage, however a trial of prazosin for nightmares is suggested. The mainstay of treatment remains psychological and Maria should be engaged with psychologist who specialises in EMDR or trauma focussed CBT.
Dr Chow’s report, and a report dated 29 April 2024 from Dr Leonard Lee, consultant forensic psychiatrist instructed by the appellant, were tendered before the primary judge. Neither psychiatrist gave oral evidence.
Dr Chow’s report included a detailed history of the respondent’s symptoms since the attack, including that she had been trying to work three days a week but struggled with that, often staying in the studio for a few hours at a time then going home. He also noted that the respondent told him that she had a past history of depression and anxiety, having taken antidepressants three times, for six months’ duration on each occasion, within the last 20 years “during life stressors”. He set out his findings on mental state examination, including that the respondent’s affect was restricted and that she reported “ongoing sleeping disturbance, flashbacks, nightmares, avoidance behaviours, triggers of psychological symptoms, depressed mood, anxiety, poor appetite, reduced interest in hobbies or activities, fluctuating motivation and energy with feeling of guilt”. His opinion was that the respondent suffered from a “psychiatric injury of trauma nature” caused by the attack and that she continued to suffer significant psychological symptoms from the attack which warranted the diagnosis of PTSD. He considered that she needed further psychological and psychiatric treatment and her condition was likely to remain ongoing. He also said that the respondent’s ability to work had been affected by the injury and that her partial incapacity was likely to remain ongoing.
Dr Lee also took a detailed history from the respondent and reported his findings on mental state examination, including that the respondent’s anxiety seemed to settle with an empathic approach. He said that on clinical examination the respondent “presented as extremely anxious and disabled”. He administered the Structured Inventory of Malingered Symptomatology (SIMS) which he described as being guided by the idea that naïve respondents are “likely to endorse bizarre, rare, atypical, or extreme symptoms on a questionnaire when they attempt to feign or exaggerate symptoms”. He explained that the SIMS generates heightened scores in groups that are known to have a raised prevalence of feigning. He said that the respondent’s score on the SIMS indicated “diagnostic certainty of feigned psychopathy”. He also administered the Memory Complaints Inventory (MCI) which he described as a “symptom validity test”. He said that the respondent’s score on the MCI indicated that it was “very likely that symptom exaggeration [was] present”. He also said that a further inventory that he administered, the Minnesota Multiphasic Personality Inventory 2 revised format, indicated that the respondent’s symptoms were “markedly exaggerated and non-credible”.
Dr Lee referred to the record of the respondent’s attendance on Dr Rehan on 25 April 2020. He makes no mention of Dr Rehan’s note that the reason for the visit was depression/anxiety. He said:
[A]s her GP records do not reveal traumatic stress just after the index incident it is difficult to reject the proposition that the index incident has not had a direct impact on her mental health.
Dr Lee made no mention of the respondent’s medical records from Highpoint Medical Centre. This is a significant omission given the matters summarised at [35]-[44] above. Dr Lee attached significance to the fact that the respondent presented to the MedFirst Medical Centre on 16 September 2020 complaining of an ant bite and bruises in her finger “with no mention of the dog bite”. He commented that it was “remarkable that she did not appear to be traumatized five months after the incident. Delayed onset PTSD is implausible”. In fact, however, in addition to the attendance on 25 April 2020 the respondent was diagnosed with “situational crisis and anxiety” on 30 April 2020, was referred to a psychologist on account of this, and on 7 May 2020 saw Dr Fu who diagnosed Acute Stress Disorder caused by the attack.
Dr Lee referred to Dr Wong’s referral to Dr Gulati on 6 February 2023. As to this, Dr Lee commented that the DASS-21 “is a sel[f-]report instrument that does not detect exaggeration”.
Dr Lee summarised and commented on Dr Chow’s report, including observing that the respondent’s “failure on symptom validity testing means that the accuracy of her history and reported symptoms is uncertain”. Curiously, given the content of the respondent’s medical records, Dr Lee commented that:
Dr Chow is unaware that there is no verification of PTSD symptomatology until Dr Wong’s referral almost three years after the incident, and does not conduct symptom validity testing.
Liability under s 25 of the CAA was not in dispute before the primary judge: [10].
As to causation, the primary judge made three key findings. First, that the respondent suffered a genuine psychiatric condition as a result of the attack: [43]. In reaching this finding, the primary judge relied upon the “nature of the attack, the contemporaneous complaints of psychological trauma, the prescription of psychotropic medication, the referral to a psychologist early on, and the treating histories referencing psychological problems following on from the attack”: [43]. The primary judge observed that there was “a conspicuous absence of any real clinical observations of the [respondent] by Dr Lee, whose conclusions are firmly rooted in the results of the testing he undertook”: [44], and that “the opinion of Dr Chow [was] more consistent with the treating clinical records and … share[d] a symmetry with the opinions expressed by the treating psychiatrist, Dr Gulati”: [45]. This led the primary judge to reject the opinion of Dr Lee that the respondent was “feigning”, noting that “[t]hat opinion was based predominantly, if not entirely, on symptom validity and other testing which, as Dr Lee’s report acknowledges, is not definitive”: [46]. His Honour added at [46]:
The clinical observations of Dr Chow and Dr Gulati supporting the [respondent] as having a genuine psychological injury are more consistent with the impressions I gained of her in the witness box, supported as she was by Mr Child.
The primary judge found that the attack caused the respondent to suffer PTSD against the background of a prior history of anxiety and depression: [51]. The primary judge characterised Dr Lee’s views as to causation as “out of step with the evidence”: [66].
Second, the primary judge held that, as liability under s 25 of the CAA is not liability “in negligence”, the issue of causation was to be determined according to the common law and not pursuant to s 5D of the CLA: [57]-[59]. His Honour observed, however, that this was of little moment as the two approaches would not yield a different outcome on the facts of this case: [59].
Third, the primary judge addressed the issue of “disentangling” the psychiatric injury caused by the attack on the respondent from the psychiatric injury associated with the injuries to Lexi. This was because the primary judge interpreted s 25(1)(a) of the CAA as conferring a right to damages only for physical and psychiatric injury consequent upon an attack on “a person”: [60], and the respondent does not now seek to challenge his Honour’s construction. As to this, the primary judge found that “the attack upon the [respondent] and the attack upon Lexi are inextricably linked”: [72]. His Honour found that there was no evidence before him which would enable “a reasoned evaluation of the psychological effects of the trauma associated with the attack on Lexi, independently of the attack on the [respondent]”, nor any reliable evidence “to dilute the causal link between the attack on the [respondent] and the PTSD [his Honour was] satisfied she suffered as a result”: [75]-[78]. His Honour held that, in these circumstances, the appellant had not discharged the evidentiary burden on her to show that part of the respondent’s injury was traceable to causes other than the attack on her: Seltsam Pty Ltd v Ghaleb [2005] NSWCA 208 at [104]-[105] (Ipp JA).
Underlying the primary judge’s assessment of damages was his Honour’s finding that the respondent was a credible witness who had suffered a genuine psychiatric injury but that her complaints were “in some respects potentially unreliable”: [47], [82], [106]. His Honour attributed this to her preoccupation with the attack and the emotional turmoil it caused: [82]. He also considered that there was cause to expect some improvement in her symptoms with the intervention of counselling, medication and the passage of time: [82]. Both of these matters led to some caution when his Honour was considering the respondent’s claim for damages for loss of earning capacity: [106].
The primary judge’s assessment of the respondent’s non-economic loss of $49,500 was based upon his Honour’s evaluation that she was 25% of a most extreme case: [83].
The $2,288.63 awarded to the respondent for past out of pocket expenses was agreed by the appellant at trial and allowed by his Honour: [85], [88].
Neither party directly addressed the standard of appellate review applicable to the appellant’s challenges to the primary judge’s assessments of non-economic and economic loss. As to this, in Harper v Bangalow Motors Pty Ltd [1990] NSWCA 85, Clarke JA explained that the same standard of appellate review will not automatically be applied to all elements, economic and non-economic, of a claim for damages. His Honour held that there was no reason why the rule in Warren v Coombes (1979) 142 CLR 531; [1979] HCA 9 should not apply to essential factual determinations whereas an approach akin to that in House v The King (1936) 55 CLR 499; [1936] HCA 40 (which his Honour referred to by reference to the decision in Miller v Jennings (1954) 92 CLR 190; [1954] HCA 65) should apply where “the decision reflects a degree of judicial prophesy or speculation”. In Diamond v Simpson (No 1) [2003] NSWCA 67 at [15], the Court described Clarke JA in Harper as having “accurately and conveniently expressed” the principles which govern the task of an appellate court in determining an appeal from an assessment of damages in personal injury cases. Having regard to Harper and Diamond, it is appropriate to identify the standard of appellate review on a challenge to an assessment of damages by reference to the character of the particular finding being challenged.
As regards the appellant’s challenges to findings of fact, such as the primary judge’s finding that the respondent genuinely suffered from ongoing psychological symptoms and that these impacted her earning capacity, the proper standard of appellate review is the correctness standard, albeit informed by the principles in Fox v Percy (2003) 214 CLR 118; [2003] HCA 22 at [28]-[29] and Lee v Lee (2019) 266 CLR 129; [2019] HCA 28 at [55]. The challenge to the primary judge’s finding of causation similarly attracts a correctness standard of appellate review, as it reflects an enquiry to which there is only one legally correct outcome: GLJ v Trustees of the Roman Catholic Church for the Diocese of Lismore (2023) 280 CLR 442; [2023] HCA 32 at [16].
As regards an assessment of non-economic loss, as was explained by Griffiths AJA in Stanberg v State of New South Wales [2025] NSWCA 127 at [109]-[110] (Mitchelmore and McHugh JJA agreeing) there may be some tension between the statement by Leeming JA in Hall v State of New South Wales [2014] NSWCA 154 at [31]-[32] and the more recent analysis of Gleeson and White JJA in White v Redding (2019) 99 NSWLR 605; [2019] NSWCA 152 at [61] and [96] respectively. In the former, Leeming JA held that “[o]rdinary principles of appellate review” apply to a finding of fact as to the severity of non-economic loss by reference to the proportion of a most extreme case. In the latter, Gleeson JA, having referred to Leeming JA’s analysis in said that a finding as to the severity of non-economic loss “does not call for a unique outcome” and thus the test for appellate intervention reflects that described in . White JA similarly held that such a decision attracts the standard of appellate review.
The appellant’s ultimate contention on ground one is that the primary judge erred in preferring the opinion of Dr Chow to that of Dr Lee. She submits that Dr Chow’s report failed to expose the reasoning linking his specialised knowledge to the conclusion reached, or to consider alternative causes of the appellant’s symptoms, and thus failed to comply with the requirements of s 79 of the Evidence Act 1995 (NSW) and as set out by Heydon JA in Makita (Australia) Pty Ltd v Sprowles (2001) 52 NSWLR 705; [2001] NSWCA 305 at [67]. She also contends that Dr Chow’s opinion is undermined by his reliance upon the respondent’s self-reported history and symptoms and the respondent’s results on the DASS-21 assessment tool. By contrast, the appellant contends, Dr Lee conducted symptom validity testing and explained how his specialised knowledge was applied to the facts.
Those contentions should be rejected. The primary judge was correct to prefer Dr Chow’s opinion over that of Dr Lee. The primary judge was also correct to describe, at [44], Dr Lee’s opinion that the respondent was feigning her symptoms as an opinion that was predicated more on his testing for malingering and feigning of symptoms than upon his own examination and assessment of the respondent (as noted above at [55]). Further, as the primary judge observed at [38] and [41]-[42], the respondent’s medical records significantly undermine Dr Lee’s observations that the respondent made no complaints of symptoms consistent with PTSD for some time after the attack.
As the primary judge correctly observed at [43], “[t]he nature of the attack, the contemporaneous complaints of psychological trauma, the prescription of psychotropic medication, the referral to a psychologist early on, and the treating histories referencing psychological problems following on from the attack” all supported a finding that the respondent suffered a genuine psychiatric condition as a result of the attack. The primary judge also relied in this regard upon his own impression of the respondent in the witness box (as supported by Mr Child’s evidence), and in this way it is apparent that the primary judge had considerable advantages over this Court.
The appellant’s criticisms of Dr Chow’s report should also be rejected. Dr Chow’s expertise was not questioned either before the primary judge or before this Court. He set out what he regarded to be relevant history and his findings on mental state examination. He referred to the respondent’s history of depression and anxiety which at times required medication. He incorporated relevant extracts from the respondent’s medical records into his summary of the background to his opinion. On a natural reading of Dr Chow’s report, his summary and opinion drew upon those matters in reasoning to an assessment of the veracity of the respondent’s symptoms and to his opinion as to diagnosis and prognosis. Both the facts relied upon and the process of reasoning leading to his opinion were sufficiently disclosed: [2005] NSWCA 152; 218 ALR 764 at [105] (Spigelman CJ, Giles and Ipp JJA agreeing). If the primary judge had not agreed that the respondent’s symptoms were genuine and were caused by the attack, having regard to all of the evidence, then Dr Chow’s opinions may have been of little or no value: (1961) 108 CLR 642 at 649 (Dixon CJ, McTiernan, Kitto, Taylor and Windeyer JJ); [1961] HCA 65.
The appellant’s overarching contention in ground two is that the primary judge failed to perform any analysis separating the effects of the attack from other significant stressors in the respondent’s life, both before and after the attack. She submits that the primary judge erroneously proceeded largely on the basis of temporal association and that this materially affected the outcome at trial.
In this regard, the appellant contends that s 5D of the CLA was applicable and required the primary judge to determine whether the attack “was the only necessary condition of the harm” and whether responsibility for the harm should be imposed. She submits that even if s 5D did not apply, the primary judge should have undertaken a structured causation analysis and distinguished the attack from other contemporaneous stressors. She identifies a range of matters which she contends were capable of producing the respondent’s psychological symptoms independently of the attack, including COVID-19 lockdowns, the suspension of weddings due to the lockdowns which affected the respondent’s business, the respondent’s history of anxiety and depression, the breakdown of the respondent’s prior intimate relationship in 2016 against a background of serious allegations against her former partner in 2012 (involving nieces of the respondent) and the death of the respondent’s father in November 2024. She contends these competing causes should have been, but were not, evaluated by the primary judge.
The appellant’s contentions begin from the erroneous starting point that s 5D of the CLA applies in the context of the claim under s 25 of the CAA which the primary judge upheld against her. Leeming JA (Macfarlan JA and Sackville AJA agreeing) held in Simon v Condran (2013) 85 NSWLR 768; [2013] NSWCA 388 at [28] that liability under s 25 of the CAA is not regulated by Pt 1A of the CLA (within which s 5D is found) as it is not dependent upon any breach of duty on the part of the dog owner. As a matter of substance, however, this does not undermine the appellant’s contention that the determination of causation involves both a question of historical fact (the “but for” question) and a normative question as to whether legal responsibility for a particular harm occurring should be attributed to a particular person: Wallace v Kam (2013) 250 CLR 375; [2013] HCA 19 at [11], [16]. In this regard, the primary judge was correct to reject the appellant’s contention below (repeated in this Court) that s 5D of the CLA applies. It should also be observed in this regard that in Wallace v Kam it was held at [16] that factual causation (which was the causation question in issue on the appeal) under s 5D(1)(a) of the CLA involves “nothing more or less than” the application of a ‘but for’ test of causation.
Under s 25(1)(a) of the an owner’s liability is in respect of “bodily injury to a person the dog wounding or attacking that person” (emphasis added). The approach to causation under s 25(1)(a) should be determined by reference to the text and purpose of the : (2016) 258 CLR 467; [2016] HCA 43 at [42]; see also (2024) 114 NSWLR 1; [2024] NSWCA 32 at [105]. I am satisfied that, for the purpose of the issues of factual causation arising in this case, the test under s 25(1)(a) should be informed by the common law principles of causation*.* Thus, whilst the primary judge erred in holding that the common law rather than the statutory test for causation under s 25(1)(a) of the applied, that error is not material as, in the context of this case, the primary judge applied the correct principles to resolve the issues of causation arising.
In ground three, the appellant contends that the evidence before the primary judge did not support a finding that the respondent had suffered impairment of her earning capacity by reason of her psychiatric symptoms, “as opposed to experiencing subjective difficulty or reduced preference for working longer hours”. In support of this contention the appellant relies upon the respondent’s continuation of substantive work activities in the period after COVID-19 restrictions were lifted, the increasing revenue generated by her business over this period in circumstances where the respondent was the only substantive contributor to the business, the respondent’s continuing ability to go to the gym after the attack and the lack of any reasoned analysis in Dr Chow’s report in support of his conclusion that the respondent “continues to struggle and [is] only able to remain working partially”. She submits that it was not open to his Honour to make an award of damages as a buffer for loss of earning capacity in these circumstances.
Ultimately, as noted above, the primary judge accepted some, but not all, of the respondent’s evidence as to her ongoing impairment of earning capacity. That finding was clearly influenced by the primary judge’s assessment of the oral evidence of both the respondent and Mr Child, albeit that that evidence was assessed in the context of the medical and financial records and the expert evidence. Whilst demeanour was not adverted to by the primary judge as a factor in his evaluation of the evidence, it is inevitable that his impression of the witnesses and his immersion in the milieu of the trial played a part in his assessment of the evidence: J and E Vella Pty Ltd v Hobson [2023] NSWCA 234 at [214] (Stern JA, Mitchelmore JA agreeing). Given this, the primary judge’s finding that the respondent’s earning capacity was impaired is one that attracts the principles of appellate restraint set out in Fox v Percy and Lee v Lee. As this finding is neither glaringly improbable nor contrary to the established facts, the appellant’s contention of error should be rejected.
In any event, the primary judge’s finding is amply supported by the respondent’s oral evidence, supported as it was by Mr Child, and in a general sense by the medical records noted above. It is also supported by Dr Chow’s opinion as to causation.
Contrary to the appellant’s submission, the primary judge was not required to identify specific functions which the respondent could not undertake in circumstances where her impairment was consequent upon a more general inability on occasion to leave the house, her need for the support of Mr Child when doing so, anxiety, lack of motivation and energy, and depression.
Ground three should be dismissed.
In large measure, the appellant’s contentions of error in the assessment of damages rely upon her contentions in grounds one to three. I have already explained why those contentions do not lead to a conclusion of error. In particular, the appellant’s contentions that the primary judge erred in his assessments of non-economic loss or in his allowance for future treatment costs were consequential upon her contention that his Honour erred in his findings as to causation and earning capacity.
The appellant also contends that the awards of damages were manifestly excessive in circumstances where the primary judge acknowledged uncertainty regarding the respondent’s financial position and that the available financial records were incomplete.
The evidence before the primary judge included personal tax records from 1 July 2014 to 30 June 2019, company taxation records from 1 July 2014 to 30 June 2023, and Business Activity Statements from July 2017 to March 2024. As the appellant submits, the primary judge acknowledged that there was uncertainty as to whether, and to what extent, the respondent had suffered economic loss having regard to the documents available and the concession made by the respondent that she was earning more per month in 2025 than she had earned per month in 2019 and that gross receipts were a gauge for how busy the business was: [113], [114], [119]. In part the uncertainty was due to gaps in the available financial records and in part it was due to the complexities of factoring in the effects of the pandemic: [117]. Notwithstanding that uncertainty, however, given the primary judge’s finding that the respondent’s earning capacity was impaired, there is no error in the primary judge having concluded on the balance of probabilities that that impairment of earning capacity led to financial loss both in the past and ongoing.
Once that is accepted, the issue raised by ground four is whether the primary judge erred in his assessment that the respondent’s impairment of earning capacity should be reflected in the award of a buffer in the sum of $35,000 or otherwise erred in his assessment of non-economic loss or in his allowance for future treatment. As to the buffer, irrespective of whether the question is one of correctness, or whether the award was within a reasonable range: see above at [65]-[69], I am not satisfied that the primary judge’s assessment was erroneous. The respondent was born in October 1970 and said that she had anticipated working until she was 70 or 80 years old. The attack occurred when she was aged 49. That resulted in about 20 to 30 years of impaired earning capacity, although the COVID-19 restrictions impacted for probably one to two of those years. In that context, the primary judge’s award of $35,000 to reflect both past and future loss of earning capacity is, if anything, conservative. It more than adequately gave effect to the uncertainties which he identified. It also reflected the fact that the respondent’s business was clearly successful and capable of generating significant income and might have been even more successful if the respondent had not suffered PTSD by reason of the attack. Nor did his Honour err in his assessment of non-economic loss or in his allowance for the costs of future treatment.
It follows that the appeal should be dismissed. Neither party contended other than that costs should follow the event.
The orders I propose are:
Appeal dismissed.
Appellant to pay the respondent’s costs of the appeal.
**********
Carretero
Ioannidis (No 3)
[2003] HCA 22
[1936] HCA 40
(2000) 50 NSWLR 388
(2016) 258 CLR 467
(2024) 114 NSWLR 1
(2003) 214 CLR 118
(2023) 280 CLR 442
(1936) 55 CLR 499
(2019) 266 CLR 129
(2001) 52 NSWLR 705
(1954) 92 CLR 190
(2018) 264 CLR 541
(1961) 108 CLR 642
(2013) 85 NSWLR 768
(2013) 250 CLR 375
(1979) 142 CLR 531
(2019) 99 NSWLR 605
(2024) 98 ALJR 1119
(2019) 99 NSWLR 447
Appeal dismissed. The appellant was ordered to pay the respondent’s costs of the appeal. The appellant was also ordered to pay the costs of the fresh-evidence application, but those costs did not include costs of or consequential upon the respondent’s subpoenas identified at [22]-[23]. [21]-[24], [96]-[97]
For the reasons set out below, the appeal should be dismissed with costs.
The appellant submitted that this evidence was apparently credible and demonstrated that the respondent had engaged in sustained professional activity and had functional capacity that was inconsistent with both her evidence at trial and the primary judge’s findings. She submitted that this evidence showed that the respondent travelled repeatedly to Melbourne for professional purposes associated with Infamous Swim, engaged in sustained on-site work in Melbourne and presented as socially engaged, confident, directive and professionally groomed in the videos posted on Instagram. She also submitted that this evidence disclosed that the respondent had negotiated a deal with the CEO of Infamous Swim, which presented a different picture to that presented by the respondent at trial. She contended that this evidence would have led to a different assessment by the primary judge of the respondent’s credibility the weight to be given to the expert evidence upon which the respondent relied and on the issues of causation and damages.
The appellant contended that this evidence was not available with reasonable diligence at the hearing before the primary judge. She said that this was because she was medically advised to minimise litigation involvement during pregnancy and the postpartum period and was not in a position to monitor social media or news publications at that time, the respondent did not disclose her involvement with Infamous Swim and, at the time of trial, she was managing full-time work and the sale of her home to fund her legal costs and relied upon her partner and legal representatives to manage communications. She submitted that:
Second, the material was not of a character that, if adduced, it would likely have had an important influence on the result of the case. The respondent was candid in her evidence at trial that she was still actively in business and it was clear that, at that time, her business (in the fashion industry) remained active and remunerative. The fact that the respondent was one of a number of investors in Infamous Swim was not inconsistent with the respondent’s evidence before the primary judge that went to her inability to work five days a week at her business. The respondent also gave evidence at trial that the nature of her business was such that she had to post frequently on social media as she was conscious of the importance of media exposure to her business. The depiction of the respondent in the Instagram posts that the appellant sought to adduce was consistent with that evidence. Even if it were accepted that the videos posted on Instagram were recorded at Infamous Swim’s premises, and thus that the material now sought to be adduced showed that the respondent travelled to Melbourne on at least one occasion (noting that it is not apparent whether the videos were filmed on one, or more, occasion), that would not undermine the credibility of the respondent’s evidence that she could no longer work full time nor would it diminish the weight that could be given to the report dated 24 February 2023 of Dr Frank Chow, psychiatrist, on which the respondent relied. Nor would any of the material sought to be adduced undermine the primary judge’s finding that the respondent had some impairment to her earning capacity by reason of her injuries (albeit less than she claimed).
It was also of some significance that (as already noted) the primary judge found that there was some unreliability in the respondent’s evidence, albeit that his Honour did not think that that was a product of a conscious attempt to mislead. His Honour found that the respondent probably had greater earning capacity, and did work more, than her evidence suggested. In these circumstances, the material sought to be adduced would not have materially impacted the primary judge’s assessment of damages.
Given this conclusion, the appellant should bear the costs of the application to adduce fresh evidence.
Several steps were taken by the respondent apparently seeking to counter the application to adduce fresh evidence. The respondent issued subpoenas to Veritas Pty Ltd, Family Services Australia and TasWomen, and on 16 March 2026 Free JA abridged the time for service of those subpoenas. His Honour ordered that the costs of that motion be costs in the appeal.
The respondent also issued subpoenas to attend to give evidence to Mr Short, Puja Payal of Veritas Law Firm and Gemma Crowe of Infamous Swim. This was a disproportionate response to the appellant’s motion. No order should be made for the costs of, or associated with, these subpoenas.
For the avoidance of doubt, the costs of the application to adduce fresh evidence (which the appellant is to bear) do not include any costs of, or consequential upon, the issuing of the subpoenas at [22]-[23] above.
Mr Child said that at the time of trial he did things to assist the respondent with her work, such as driving her to her work in Balmain and staying there for five to six hours while she works on a Saturday and organising to get her to work or get her home if she had to work on a Friday. He said that when the respondent worked from home, he made her appointments and her business emails went to his personal inbox.
Financial records tendered at trial, and on which the respondent was cross-examined, showed that the respondent’s business was more successful in the period between 2022 and 2024 than it had been prior to COVID-19 restrictions. In 2019 her total sales were $106,440, but in 2022, 2023 and 2024 they were, respectively, $164,925, $147,014 and $142,845. When it was put to her in cross-examination that this showed significant increases in total sales since the attack she said:
The award of a $15,000 buffer for future treatment expenses was predicated upon the respondent’s need for future medication and the possibility of her availing herself of psychiatric or psychological treatment in the future: [99].
In awarding a $35,000 buffer for past and future loss of earning capacity the primary judge noted that the respondent’s evidence as to the extent of her incapacity sat uncomfortably with the frequency of her gym attendance, her failure to avail herself of treatment, the substantial income her business had generated since the COVID-19 pandemic, Dr Wong’s clinical note that the respondent “goes to gym, work without trouble”, and Mr Child’s evidence that he would wait five or six hours for the respondent at her work on a Saturday: [110], [119], [121], [122]. His Honour also observed that the respondent’s pre-existing psychological problems may have “come against her in any event”: [124]. He found that the respondent “probably does more work than her evidence suggests”: [123]. Notwithstanding these matters, however, the primary judge was satisfied that the respondent had suffered some loss of earning capacity as a result of the psychological injury caused by the attack, albeit that the extent of that loss was uncertain: [118], [124].
No award was made for domestic assistance: [137].
As Gageler J explained in Minister for Immigration v SZVFW (2018) 264 CLR 541; [2018] HCA 30 at [49], and has subsequently been confirmed in GLJ at [15] (Kiefel CJ, Gageler and Jagot JJ) and Moore (a pseudonym) v The King [2024] HCA 30; 98 ALJR 1119 at [15] (Gageler CJ, Edelman, Steward, Gleeson and Beech-Jones JJ), the correctness standard of review is apt where the legal criterion applied demand a unique answer but the House v The King standard applies where the conclusion tolerates a range of outcomes, albeit that the line between the two is “not bright”. Consistent with the careful and persuasive analysis of Gleeson and White JJA in White, the characterisation of a plaintiff’s injuries by reference to a proportion of a most extreme case, so as to assess damages for non-economic loss, is a necessarily imprecise and impressionistic exercise. It would be wholly artificial to treat such an exercise as one in which there is only one legally correct answer or a unique outcome (see by analogy, Motor Accident Commission v Raccanello [2025] SASCA 146 at [123]-[125] (Livesey P) and [1117] (S Doyle JA) and Transport Accident Commission v Pastuch [2025] SASCA 120 at [44], (S Doyle JA, Livesey P and Bleby JA agreeing)). In these circumstances, the House v The King standard of appellate review should apply to challenges where it is alleged that the primary judge has erred in that characterisation. Having said that, however, I am satisfied that the primary judge’s assessment of non-economic loss was correct here, so my conclusion does not depend upon which standard of appellate review applies.
As regards the appellant’s challenge to the primary judge’s award for loss of earning capacity, there is a line of authority that indicates that the question for the appellate court is whether the amount awarded was within the range available to the trial judge or within “a reasonable range”: Loxton v State of New South Wales [2002] NSWCA 194 at [46] (Sheller JA, Campbell AJA agreeing); Pollard v Baulderstone Hornibrook Engineering Pty Ltd [2008] NSWCA 99 at [82] (McColl JA, Mason P and Beazley JA agreeing); Chen by her tutor Huang v Kmart Australia Ltd [2023] NSWCA 96 at [51] (White JA, Griffiths AJA and Weinstein J agreeing); see also Langdon v Carnival PLC [2024] NSWCA 168 at [163] (Ward P, Kirk JA and Basten AJA agreeing). This reflects the “necessarily impressionistic” nature of the exercise: Brear v James Hardie & Co**y Pty Ltd (2000) 50 NSWLR 388; [2000] NSWCA 352 at [49] (Mason P, Spigelman CJ and Priestley JA agreeing), which is “imprecise and indeterminate” with wide parameters: Loxton at [41]. However, again, even if the correctness standard were to apply, it would make no difference to the outcome of this appeal.
Dr Chow’s opinions as to diagnosis and causation implicitly reject any suggestion that the respondent was feigning her symptoms or that pre-existing factors caused her current symptoms. He did not need to detail his process of reasoning as it is readily apparent from both the structure and content of his report. Nor did he need to administer symptom validity testing in order to support those opinions. It is well within the expertise of a consultant psychiatrist to carry out assessments (whether therapeutic or medico-legal) on the basis of the account given by the person being assessed and to form a view as to the extent to which the person being assessed genuinely suffers from the symptoms or impairment they describe. That, together with mental state assessment, may be an important element of psychiatric assessments and decisions to recommend or prescribe medication or other treatment, with a view to avoiding any unnecessary administration of costly and potentially harmful medication. Whilst the psychiatrist’s assessment in no way binds the Court, it is not irrelevant to the Court’s own assessment as to whether or not symptoms described by a plaintiff (or here by the respondent) are genuinely experienced or reflect a psychiatric condition.
Ground one should be dismissed.
The primary judge considered the issue of factual causation first by asking whether the attack caused the respondent to suffer both from a psychiatric condition and from ongoing symptoms of that condition (above at [56]). The medical records that the primary judge relied upon to reach the conclusion that it did strongly support his Honour’s conclusion. Those records disclose initial anxiety related to the attack and then a series of attendances on general practitioners, a psychologist and a psychiatrist, in which the development of her symptoms since the attack, her symptoms reflecting anxiety and a stress disorder or situational crisis and the link between her symptoms and the attack, were noted (albeit that not all of these matters are identified in each of the clinical notes): Dr Rehan on 25 April 2020, Dr Habashi on 30 April 2020, Mr Fu on 7 May 2020, Dr Habashi on 9 July 2020, Dr Wang on 24June 2021, Dr Liu on 27 October 2021, Dr Tay on 3 March 2022, Dr Wong on 14 March 2022, Dr Wong on 7 September 2022, Dr Wong on 7 October 2022, Dr Wong on 6 February 2023 and Dr Gulati on 14 February 2023 (summarised above at [34]-[45]). The opinion of Dr Chow also supports the primary judge’s assessment.
The primary judge’s analysis of causation is also supported by the absence of any suggestion in the medical records that the respondent’s symptoms were consequent upon any of the events which the appellant posits as possible alternative causes or her psychological symptoms or impairment. It is also supported by the respondent’s oral evidence (which is supported by Mr Child’s oral evidence) that her previous anxiety and stress was “very different” to what happened after the attack and that her anxiety as at the time of trial was also very different to how it had been before the attack. Further, it was not put to the respondent in cross-examination that she had suffered symptoms of stress and anxiety by reason of the matters that the appellant now contends the primary judge should have considered as alternative causes or contributors to her psychological symptoms after the attack.
In this context, the primary judge’s analysis of the issue of causation was both adequate and correct. It was not incumbent upon the primary judge to scrutinise the material before him to ascertain whether there might have been events in the respondent’s life, other than the attack, which could potentially have led to anxiety or a stress reaction in circumstances where these were neither put to the respondent and neither of the experts suggested that any of the posited competing causes had caused or contributed to the respondent’s PTSD (although Dr Lee did suggest, as a matter of generality and not with respect to the respondent, that one thing which can result in misreporting and simulation of clinical disorders is a dysfunctional life and he referred in his report to the impact of the COVID-19 pandemic on her business in the context of his opinion that it was impossible to reject the possibility of malingering).
Ground two should be dismissed.
Ground four should be dismissed.