ADLAM v ACCIDENT COMPENSATION CORPORATION [2017] NZCA 457

ADLAM v ACCIDENT COMPENSATION CORPORATION [2017] NZCA 457 [16 October 2017]IN THE COURT OF APPEAL OF NEW ZEALANDCA442/2016[2017] NZCA 457BETWEEN JEREMY CAMERON ADLAMAppellantAND ACCIDENT COMPENSATIONCORPORATIONRespondentHearing: 6 April 2017Court: French, Cooper and Asher JJCounsel: H G de Groot for AppellantA S Butler and A F Pullar for RespondentJudgment: 16 October 2017 at 3 pmJUDGMENT OF THE COURTA The application to adduce further evidence is granted.B We answer question (a) of the case stated as follows:Was the High Court correct to find that a failure to treat cannot occur incircumstances where there are no indications for a different treatmentcourse?Yes.C The appeal is dismissed.___________________________________________________________________REASONS OF THE COURT(Given by Cooper J)Introduction[1] Section 20(1) of the Accident Compensation Act 2001 (the Act) provides coverunder the Act for personal injury suffered in New Zealand on or after 1 April 2002.To be covered, the injury sustained must be a personal injury of a defined kind (here,a physical injury)1 and the injury must be described in one of the paragraphs of s 20(2).[2] Section 20(2)(b) refers to personal injury that is "treatment injury" suffered bythe person. This appeal requires the Court to consider what is meant by the term"treatment injury". That necessitates an examination of both the definition of"treatment injury" in s 32 of the Act and the definition of "treatment" in s 33, whichincludes a failure to provide treatment or to provide it in a timely manner.2[3] The case comes to this Court as an appeal by way of case stated on questionsof law under s 163(1) of the Act. The issues have earlier been considered by the HighCourt,3 and the District Court.4 The High Court allowed an appeal by the AccidentCompensation Corporation (ACC), overturning the District Court's decision that therewas cover under the Act for treatment injury. Before turning to the questions and therelevant statutory provisions we set out the facts as recited in the case stated by theHigh Court.Facts[4] The facts stated by Gendall J are as follows:(a) The appellant, Jeremy Adlam, is now aged 16 years. He was born on20 March 2000. Immediately prior to his birth by emergencycaesarean section he suffered a neurological injury described as intra-partum hypoxia which later caused severe cerebral palsy. Prior to thatevent, the appellant was a healthy, viable child.(b) The appellant suffers marked neurocognitive disability which affectsevery area of his daily living, including deficits in terms of mobility,upper extremity function, swallowing and fatigue. His mother acts ashis full-time carer.1 Accident Compensation Act 2001, s 26(1)(b).2 Accident Compensation Act, s 33(1)(d).3 Accident Compensation Corporation v Adlam [2016] NZHC 1487, [2016] 3 NZLR 497.4 Adlam v Accident Compensation Corporation [2015] NZACC 230. The District Court heard anappeal under s 149 of the Act against a review decision.(c) During her pregnancy with the appellant, his mother, Mrs KymAdlam, was hospitalised at 27 weeks gestation as a result ofspontaneous and premature rupture of her membranes. She wastransferred as an in-patient to National Women's Hospital in Aucklandwhere she was seen by a paediatric registrar and her condition wasstabilised. From that point on, her pregnancy and the foetal conditionwere closely monitored.(d) Over the next four weeks, Mrs Adlam remained at National Women'sHospital. During this time, she appeared to be clinically well and wasreacting appropriately to cardiotocography (CTG) tracing. There wasno evidence of uterine activity, chorioamnionitis, or pre-term labour.She was transferred to New Plymouth Hospital near her home whereshe was seen by a specialist obstetrician. The plan was for her todeliver at 34 weeks gestation.(e) On 17 March 2000, a doctor at New Plymouth Hospital noted that allrecordings were normal and the foetus was in the cephalic (headdown) position.(f) On 20 March 2000, Mrs Adlam spontaneously developed contractionswith increased discharge of liquor. At first, a CTG showed goodvariation with one late deceleration. At 1.00 pm that day, Mrs Adlamwas reviewed and noted as being 0.5 cm dilated and possibly in earlylabour. The foetal heart rate was monitored over the course of theafternoon.(g) At 5.35 pm, it was noted that Mrs Adlam's temperature had increased.She was placed on IV antibiotics. Between 6.45 pm and 6.50 pm thatnight the foetal heart rate had decreased to 95–100 beats per minute.At that point a decision was made to perform an emergency caesareansection. Mrs Adlam was taken to theatre at 7.00 pm. Anaesthesia wasadministered at 7.15 pm.(h) The appellant was delivered at 7.34 pm. At birth his APGAR scoreswere six at five minutes and eight at 10 minutes and his initial bloodgases were suggestive of peripartum asphyxia. He was transferred toa neonatal unit where he was noted to have poor muscle tone,decreased reflexes, low blood pressure and poor saturation levels.(i) In the District Court, Judge MacLean found the appellant had suffereda treatment injury. In making that finding, however, Judge MacLeanreached the clear conclusion that, once it became apparent atNew Plymouth Hospital that the foetal condition was compromised,Mrs Adlam's treatment providers acted as quickly as possible. JudgeMacLean accepted that the caesarean was only appropriate when feverbecame apparent. The Judge held that "it is unrealistic to concludethat those actions could safely have been done any faster or that theywere not done in a timely way".(j) The District Court decision therefore found that the appellant wasentitled to cover despite the fact that first, there were no observableindications prior to the fever that warranted medical intervention, andsecondly, that the only basis on which treatment injury wasestablished was through objective hindsight.(k) Had a caesarean been performed earlier, it is likely that the appellantwould not have suffered personal injury.(l) There were no indications for a different treatment course in thecircumstances.(Footnote omitted.)The questions of law[5] In his substantive judgment overturning the District Court decision, Gendall Janswered two questions of law in the negative. They were:5(a) Can a "failure to treat" occur in circumstances where there are noobservable indications for a different treatment course?(b) Does a treatment injury occur where, with the benefit of objectivehindsight, a different course of treatment could have yielded a betterresult?[6] In the decision granting leave to appeal, the Judge accepted that the followingquestions of law would appropriately be asked of this Court:6(a) Was the High Court correct to find that a "failure to treat" cannot occurin circumstances where there are no indications for a differenttreatment course?(b) Was the High Court correct to find that a treatment injury does notoccur where, with the benefit of objective hindsight, a different courseof treatment could have yielded a better result?(c) Was the High Court correct to find that "some form of responsibilityfor either a failure to provide treatment or the provision of treatmentmust be established on the balance of probabilities in order for aclaimant to receive cover"?(d) Was the High Court correct to find that "determining whether aparticular treatment is clinically indicated necessarily requiresconsideration of the alternative option or options available to thehealth professional, given the state of knowledge at the time, andwhether there were contra indications telling against thosealternatives?5 Accident Compensation Corporation v Adlam, above n 3, at [8] and [64].6 Adlam v Accident Compensation Corporation [2016] NZHC 1967 at [21].[7] As Mr de Groot submitted for the appellant, a single issue underlies all fourquestions, namely what is the meaning of the words "a failure to provide treatment, orto provide treatment in a timely manner" in s 33(1)(d) of the Act. For reasons to bediscussed the appellant contends that: "failure" should be taken to mean failuresimpliciter; to establish a causative failure it need only be shown that an availabletreatment could have been given in an attempt to avoid or mitigate the personal injurysuffered; and if given the treatment would, on the balance of probabilities, haveavoided or mitigated the personal injury.[8] Whether the appellant is correct turns on an analysis of the text of the relevantstatutory provisions, considered in the light of their purpose.7 In determining purpose,regard must be had to both the immediate and general legislative context, and anyrelevant social or other objective of the enactment.8[9] The consequences of that approach to this Act was summarised by McGrath Jin a well-known passage in Harrild v Director of Proceedings:9[130] The policy of successive accident compensation statutes inNew Zealand, including the 2001 Act, has been to provide compensation forpersons suffering personal injury without requiring that they show fault toestablish their entitlement. The legislative bar to suits at common law is theconsequence of the universal nature of coverage under the legislation. A"generous, unniggardly interpretation" of what was personal injury byaccident under earlier accident compensation legislation was seen bythis Court as in keeping with that legislative policy: Accident CompensationCorporation v Mitchell [1992] 2 NZLR 436 at 438 per Richardson J. I regardthat approach to interpretation as unaffected by the narrower approach todefining personal injury since the 1992 Act .The relevant statutory provisions[10] As noted above, the combination of s 20(1) and 20(2)(b) provides cover forphysical injury that is "treatment injury". That term is defined in s 32(1) of the Act asfollows:32 Treatment injury7 Interpretation Act 1999, s 5(1).8 Commerce Commission v Fonterra Co-operative Group Ltd [2007] NZSC 36, [2007] 3 NZLR767 at [22].9 Harrild v Director of Proceedings [2003] 3 NZLR 289 (CA).(1) Treatment injury means personal injury that is—(a) suffered by a person—(i) seeking treatment from 1 or more registered healthprofessionals; or(ii) receiving treatment from, or at the direction of, 1 ormore registered health professionals; or(iii) referred to in subsection (7); and(b) caused by treatment; and(c) not a necessary part, or ordinary consequence, of thetreatment, taking into account all the circumstances of thetreatment, including—(i) the person's underlying health condition at the timeof the treatment; and(ii) the clinical knowledge at the time of the treatment.[11] In addition, s 32(2) and (3) provide:(2) Treatment injury does not include the following kinds of personalinjury:(a) personal injury that is wholly or substantially caused by aperson's underlying health condition:(b) personal injury that is solely attributable to a resourceallocation decision:(c) personal injury that is a result of a person unreasonablywithholding or delaying their consent to undergo treatment.(3) The fact that the treatment did not achieve a desired result does not,of itself, constitute treatment injury.[12] Section 33(1) defines "treatment" as follows:33 Treatment(1) For the purposes of determining whether a treatment injury hasoccurred, or when that injury occurred, treatment includes—(a) the giving of treatment:(b) a diagnosis of a person's medical condition:(c) a decision on the treatment to be provided (including adecision not to provide treatment):(d) a failure to provide treatment, or to provide treatment in atimely manner:(e) obtaining, or failing to obtain, a person's consent to undergotreatment, including any information provided to the person(or other person legally entitled to consent on their behalf ifthe person does not have legal capacity) to enable the personto make an informed decision on whether to accept treatment:(f) the provision of prophylaxis:(g) the failure of any equipment, device, or tool used as part ofthe treatment process, including the failure of any implant orprosthesis (except where the failure of the implant orprosthesis is caused by an intervening act or by fair wear andtear), whether at the time of giving treatment or subsequently:(h) the application of any support systems, including policies,processes, practices, and administrative systems, that—(i) are used by the organisation or person providing thetreatment; and(ii) directly support the treatment.[13] As we have noted, the key provision to be construed in this case is s 33(1)(d).On the facts set out in the case, the issue is whether there was a failure to providetreatment in circumstances where a caesarean section could have been performed atan earlier point, but it was in fact not then indicated. The question turns on whether"failure" in this context should be construed as meaning not doing something thatshould have been done or whether it means simply not doing something that couldhave been done. The distinction is exemplified by different meanings of "failure"given in The Oxford English Dictionary: "[a] failing to occur, be performed, or beproduced" and "an omitting to perform something due or required; default".10Appellant's argumentText[14] Mr de Groot's primary argument was that the word "failure" should beconstrued to mean failure simpliciter, or non-occurrence, without any connotation ofdeparture from a standard. He submitted this was the meaning to be derived from the10 JA Simpson and ESC Weiner (eds) The Oxford English Dictionary (2nd ed, Oxford UniversityPress, Oxford, 1989) at 667.plain language of s 33(1)(d) itself, as well as being consistent with other uses of theword "failure" in the statute. He contrasted s 33(1)(d) with the following provisionsin the Act that employ the word in conjunction with a statement of a relevant standard:(a) s 158(3), which refers to the "failure of an assessor to perform his orher duties", those duties being set out in s 158(1);(b) s 179(1)(e), which uses the expression "failure to reach the standardrequired by the audit", in provisions dealing with the Act's regime forthe upward adjustment of levies;(c) s 242(1), which refers to intentional failure to deliver statements,documents and other information required by s 241 of the Act (theseare statements or estimates of earning required by ACC in relation tolevies);(d) s 249(1)(a), which provides for a penalty in the case of an overpaymentmade as a result of a "failure by the person to inform the Corporation"of matters "that the person ought to have known" were relevant to thepayment;(e) s 316(2), which provides that a person may not be convicted of anoffence under subs (1)(a) (knowingly not making a deduction requiredto be made under s 221) if the person satisfies the court that the amountof the required deduction has been paid and that the "failure to makepayment of the deduction within the prescribed time was due to illness,accident, or other cause beyond the person's control"; and(f) s 344(2), which provides for the imposition of administrative sanctionson insurers "in respect of any failure to comply with the insurer'sobligations under the 1998 Act".1111 The reference is to the Accident Insurance Act 1998: Accident Compensation Act, s 341(1),definition of "1998 Act".[15] Mr de Groot also referred to other statutory examples of the word "failure"being used with respect to specified standards in the Corrections Act 2004 and theCrimes Act 1961. He sought to contrast these provisions with others where the wordfailure appeared without reference to standards, so as to connote simply thenon-occurrence of the fact. As an example, he referred to s 58(1) of the Act, whichprovides that a claimant must be regarded as having a decision by the ACC that he orshe has cover for a personal injury in respect of which a claim was made when "theCorporation fails to comply with a time limit under section 56 or section 57".[16] Mr de Groot similarly made reference to the heading of s 296, "Failure to reachagreement", and to s 317(7), which says that nothing in that section is affected by,amongst other things, "the failure or refusal of any person to lodge a claim for personalinjury" covered by the Act.12[17] Mr de Groot submitted that these provisions demonstrate that the word"failure" can have two distinct meanings in statutory provisions depending on context:simply not doing something, and a failure to comply with a standard. He argued thatif Parliament had intended s 33(1)(d) of the Act to imply a normative test, there wouldhave been a reference to some sort of standard in the provision. That had not beendone in s 33(1)(d), nor in the case of para (g), which speaks simply of the "failure ofany equipment, device, or tool".[18] Mr de Groot bolstered these arguments by reference to the definition of"failure" in The New Zealand Oxford Dictionary, focusing on two of the meaningsgiven, namely "non-performance, non-occurrence" and "breaking down or ceasing tofunction (heart failure; engine failure; computer failure)".13 He referred also to othercontexts in which "failure" has been treated as not importing any notion of fault.[19] In summary, while conceding that failure can imply departure from a standard,Mr de Groot submitted that where the word is used in the Act on its own without anyreference to a standard it should not be treated as requiring departure from a standard.12 Section 317(1) provides that there can be no damages claimed for personal injury covered by theAct brought independently of the Act.13 Tony Deverson and Graeme Kennedy (eds) The New Zealand Oxford Dictionary(Oxford University Press, Melbourne, 2008) at 381.Purpose[20] Mr de Groot supported his interpretation of "failure" by reference to the historyand purpose of the accident compensation legislation and the non-fault policyunderpinning it. He referred to the 1967 report of the Royal Commission of Inquiry,Compensation for Personal Injury in New Zealand, also known as theWoodhouse Report.14 He noted the Commission's intention that the scheme:15 provide immediate compensation without proof of fault for every injuredperson, regardless of his or her fault, and whether the accident occurred in thefactory, on the highway, or in the home.[21] When it was first enacted, the Accident Compensation Act 1972 had providedthat "cover" would exist for "personal injury by accident" but was shortly thereafteramended to make it clear that the physical or mental consequences of injury oraccident were also covered, together with "medical, surgical, dental, or first aidmisadventure". The expression "medical misadventure" was not defined, but wasdeveloped in case law to a position summarised by this Court in Childs v Hillock,which approved the following four propositions:16(1) Medical negligence or medical error is medical misadventure.(2) A totally unforeseen adverse consequence of medical treatment ismedical misadventure.(3) An adverse consequence of such treatment which is within the normalrange of medical or surgical failure attendant upon such treatment isnot medical misadventure.(4) An adverse consequence of such treatment which is outside thenormal range of medical or surgical failure attendant upon suchtreatment is medical misadventure.[22] Professor Todd regarded Childs v Hillock as confirming a two-limb approachdeveloped by the Courts, asking whether there had been either medical negligence ormedical mishap. He considered the judgment acknowledged the need to differentiatebetween patients who were covered for compensation and patients who were simplyreceiving medical treatment because they were ill. Some unexpected or14 Royal Commission of Inquiry Compensation for Personal Injury in New Zealand (Report,December 1967).15 At [18].16 Childs v Hillock [1994] 2 NZLR 65 (CA) at 72."accident-like" event was required to remove the case from the category of sicknessor disease (not covered) into the category of medical misadventure (for which therewas cover).17[23] The approach developed by the courts was reflected in the enactment of theAccident Rehabilitation and Compensation Insurance Act 1992, in which s 5 defined"medical misadventure" (as meaning personal injury resulting from medical error ormedical mishap), "medical error" and "medical mishap". "Medical error" was definedas follows:"Medical error" means the failure of a registered health professional toobserve a standard of care and skill reasonably to be expected in thecircumstances. It is not medical error solely because desired results are notachieved or because subsequent events show that different decisions mighthave produced better results.[24] "Medical mishap" was defined as an adverse consequence of treatmentproperly given if the likelihood of the adverse consequence of the treatment occurringwas rare and the adverse consequence of the treatment was severe.[25] Definitions of these terms were again found in the Accident Insurance Act 1998and the Injury Prevention, Rehabilitation, and Compensation Act 2001 (the currentAct before it was renamed), where they applied until 30 June 2005. The InjuryPrevention, Rehabilitation, and Compensation Amendment Act (No 2) 2005, in forcerelevantly from 1 July 2005, then added the provisions concerning treatment injuryand treatment that currently apply.18[26] As Mr de Groot submitted, the 2005 amendment followed a comprehensivereview of the previous statutory provisions concerning medical misadventure. Thereview was heralded by a 2003 publication, Review of ACC Medical Misadventure:Consultation Document.19 The executive summary of that document referred to thefact that people are sometimes unintentionally injured when receiving treatment. It17 Stephen Todd "Treatment Injury in New Zealand" (2011) 86 Chi-Kent LRev 1169 at 1186–1187.18 There is one exception in that a new s 34(5) dealing with claims based on medical misadventure(prior to 1 July 2005), which was added by s 4 of the Accident Compensation Amendment Act(No 2) 2013 with effect from 5 December that year. That amendment is not relevant here.19 Ruth Dyson Review of ACC Medical Misadventure: Consultation Document(Accident Compensation Corporation, 2003).recorded that the Act's provisions for medical misadventure had been reviewed bygovernment agencies and said:20The review has found that the ACC medical misadventure process generallyworks well in comparison to the few countries that have equivalent schemes,and is a superior alternative to seeking remedies through the Courts. Thereview has identified that claimants and health professionals see the currentmedical misadventure criteria to be unfair, confusing and arbitrary. To improve ACC's responsiveness to claimants and to move towards a fairersystem, the review has suggested that the current requirement to establish faultfor individual health professionals should be removed from the legislativeprovisions.[27] The explanatory note for what was to become the Injury Prevention,Rehabilitation, and Compensation Amendment Act (No 2) included the statementthat:21The cover provisions no longer distinguish between error and mishap, butsimply refer to "treatment injuries". Treatment injuries are ones that resultfrom the treatment, not the underlying condition of the patient. The provisionsdo not require ACC to make any findings of fault and, in this sense, it [sic] isconsistent with the no-fault nature of the scheme generally.[28] Mr de Groot also referred us to a statement made by the Minister moving thefirst reading of the Bill noting that the new category of treatment injury would:22 remove the requirement to find fault, as is the case with some claimants formedical misadventure, or to prove that a medical injury is rare and severe, asis the case with other claims for medical misadventure, before a patient isentitled to accident compensation cover. This approach is more consistentwith the no-fault nature of the scheme. As a result of these changes, morepeople will be eligible for cover, and outcomes for claimants will beconsiderably improved.[29] Further, in moving the second reading of the Bill, the Minister said:23No longer is there a distinction between medical error and medical mishap.Claimants do not have to demonstrate error or show that their condition is bothrare and severe. The new cover provisions provide that treatment injuries aresimply ones that result from the treatment. Punitive fault finding is alsoremoved from the processes of the Accident Compensation Corporation(ACC).20 At 2.21 Injury Prevention, Rehabilitation, and Compensation Amendment Bill (No 3) (165-1)(explanatory note) at 3.22 (5 August 2004) 619 NZPD 14695.23 (3 May 2005) 625 NZPD 20162.[30] At the hearing in this Court we granted an application by Mr De Groot toproduce and address further evidence not placed before the High Court that wassubsequently obtained under the Official Information Act 1982. The further evidenceconsisted of a number of Cabinet Social Development Committee reports and minutesgenerated as the 2005 amendment was developed.24 Mr de Groot relied on thismaterial as showing that the Government altered course by deciding not to proceedwith a hindsight exclusion as a basis for declining cover under the new treatment injuryproposal then under consideration. He submitted the further documents establishedthat government thinking evolved to a position where treatment injury requires onlyan objective hindsight finding of fact, with cover limited solely by the expressexclusions.25[31] Mr de Groot was also able to refer to the following statement in Personal Injuryin New Zealand:26It is clear from the parliamentary record in 2004–05 that — consistent withthe no-fault nature of the accident compensation scheme generally —Parliament intended there to be no requirement to find medical fault or errorbefore treatment injury cover could be granted. It follows that the word "failure" in s 33(1)(d) and (1)(g), and the words"failing to obtain" in s 33(1)(e) were intended to be interpreted in a neutralway. That is, they refer to something which has not happened or worked, forwhatever reason. Similar to where treatment has actually been given(s 33(1)(a)), there is no requirement to establish fault or error.[32] Mr de Groot also drew support from an article of Professor Manning publishedin Medical Law International, "Plus ça change, plus c'est la mȇme chose: Negligenceand treatment injury in New Zealand's accident compensation scheme".27 ProfessorManning is critical of a trend she discerns in the case law in which notions ofnegligence have crept back into the interpretation of the treatment injury criteria. Sheexpresses the view that the purpose of the 2005 reform is in danger of misfiring as a24 We nevertheless note and repeat the reservations about the use of such material that this Courtexpressed in Skycity Auckland Ltd v Gambling Commission [2007] NZCA 407, [2008] 2 NZLR182 at [39]–[42].25 Under the Act as it stood prior to the 2005 amendment, s 33(4)(b) provided that medical error didnot exist solely because subsequent events show that different decisions might have producedbetter results. An equivalent provision was not included in the 2005 amendment.26 Personal Injury in New Zealand (looseleaf ed, Thomson Reuters) at [AC33.01(3)].27 Joanna M Manning, "Plus ça change, plus c'est la mȇme chose: Negligence and treatment injuryin New Zealand's accident compensation scheme" (2014) 14 Medical Law International 22.consequence. She considers this needs to be corrected not only to keep faith with theintentions of the reform, but also to maximise provider trust and cooperation in theclaims process in the interests of claimants.ACC's argumentText[33] For ACC, Mr Butler submits that to establish a "failure to provide treatment"under s 33(1)(d), a claimant must show that an alternative treatment that would haveprevented the injury suffered both could and should have been given (in preference tothe treatment course that was taken), according to the clinical indications in thecircumstances.[34] Mr Butler argued that the words "failure to provide treatment" used ins 33(1)(d) strongly imply a departure from a standard, as Gendall J held in theHigh Court.28 He referred to the Concise Oxford English Dictionary definition offailure as, relevantly, "the omission of expected or required action: their failure tocomply with the rules".29 He submitted that the ordinary connotations of "failure",include "deficiency" and "shortcoming".30 Although "failure" can also be defined inmore neutral terms, for example as a lack of success, such an interpretation isprecluded here by s 32(3) of the Act, which provides that the fact treatment did notachieve a desired result does not of itself constitute treatment injury. On this approach,a "failure to provide treatment" should be interpreted by reference to the standards onwhich treatment decisions are based, those being what is clinically indicated.Consequently, a failure to provide treatment would only arise where some otheravailable and preferable course to that in fact taken was not provided[35] Mr Butler referred to various contextual considerations that he submittedreinforce this conclusion. Importantly, the definition of "treatment injury" in s 32(1)requires not only that the personal injury be suffered,31 but also that it be "caused by28 Accident Compensation Corporation v Adlam, above n 3, at [36].29 Angus Stevenson and Maurice White (eds) Concise Oxford English Dictionary (12th ed, OxfordUniversity Press, Oxford, 2011).30 Citing Gerry Breslin and Lorna Gilmour (eds) Collins English Thesaurus (HarperCollins,Glasgow, 2013) at 314.31 Accident Compensation Act, s 32(1)(a).treatment"32 and "not a necessary part, or ordinary consequence, of the treatment".33Mr Butler submitted that, in the case of a failure to treat, some other course wouldneed to have been clinically indicated before it could be said the failure had causedinjury.Purpose[36] Mr Butler submitted that the purpose of the 2005 amendment was to addressproblems identified with the previous medical misadventure provisions, under whichcover was available in cases of "medical error" and "medical mishap". The formerrequired proof that a registered health professional had failed to observe the standardof care and skill reasonably to be expected in the circumstances.34 "Medical mishap"by contrast, required proof of an injury that was caused by treatment where treatmenthad been given properly and the injury was rare and severe.35 At the time of theamendment ACC was obliged to report findings of "medical error" to the relevantprofessional body and to the Health and Disability Commissioner, identifying theresponsible health professional or organisation by name.36[37] Mr Butler referred to the stated purpose set out in the explanatory note tothe Bill of the 2005 amendment, which referred to reducing the impact and incidenceof medical injuries by "building a partnership with the health sector that":37• moves away from the punitive system of finding medical error andreporting medical misadventure cover decisions:• gains the co-operation of the health sector to participate in the claimprocess, to support claimants to make claims and to provide necessarymedical reports and advice to ACC promptly:• supports and acts as a catalyst for quality/learning initiatives in thehealth sector:• helps maintain the safety net through reporting cases where there is arisk of harm to the public.32 Section 32(1)(b).33 Section 32(1)(c).34 Injury Prevention, Rehabilitation, and Compensation Act 2001, s 33(1) (prior to the 2005amendment).35 Section 34(1) (prior to the 2005 amendment).36 Section 284(2) and (6) (prior to the 2005 amendment).37 Injury Prevention, Rehabilitation, and Compensation Amendment Bill (No 3) (165-1)(explanatory note) at 2.[38] Mr Butler submitted that to achieve those purposes, the 2005 amendment:(a) replaced cover for medical error and medical mishap within a singleoverarching framework of "treatment injury". This abandoned therequirement to attribute blame and for injuries caused by treatment tobe "rare" and "severe";(b) replaced ACC's reporting requirements with a broader obligation toreport a risk of harm to the public where ACC believes such a risk existsfrom information collected in processing claims;38 and(c) repealed the provisions that enabled health professionals to challengemedical misadventure decisions given that they were only necessarywhere blame had to be attributed to individual professionals in orderfor a claimant to obtain cover.[39] He submitted that the amendment did not reduce the relevance of clinicalindications to treatment injury determination, as an analysis of the text showed. Hedrew a distinction between the attribution of blame to individual professionals anddetermining whether treatment accorded with what was clinically indicated in givencircumstances. By retaining the relevance of clinical indications, Parliament allowedfor injuries caused by treatment to be assessed against known criteria for the purposesof determining cover. There was still an obligation on a claimant to show that medicalintervention had caused a particular injury although the need to attribute blame for thatinjury had been removed. The focus moved from finding fault with the conduct of thehealth professionals involved to a focus on the sufficiency of the treatment the patienthad received.Analysis[40] The drafting of ss 32 and 33 means that both must be applied together for thepurpose of ascertaining whether a person has suffered a treatment injury. A treatment38 Accident Compensation Act, s 284.injury must involve some act or omission that has a causative effect in producing thepersonal injury.[41] Once an event has occurred that can properly be considered as within one ofthe paragraphs in s 33(1), it must then be asked whether that event has given rise to apersonal injury that qualifies as a treatment injury under s 32(1). And the requirementsof s 32(1) are cumulative in the sense that all must be satisfied for there to be atreatment injury. Thus it is not enough for this to be established by pointing to apersonal injury that a person has suffered while receiving treatment. It must also beshown that the personal injury has been caused by treatment. Then it must bedemonstrated that the injury was not a necessary part or ordinary consequence of thetreatment taking into account all the circumstances in which the treatment hasoccurred.[42] The way in which a treatment injury arises can be illustrated by consideringindividual paragraphs in s 33(1). Take first the giving of treatment itself, referred toin para (a). In most cases the application of this provision will be straightforward, butit must be read together with s 32(1)(b) and (c). Assuming s 32(1)(b) is satisfied therewill be a personal injury. That then leads to the question of whether the injury sufferedwas not a necessary part or ordinary consequence of the treatment, taking into accountthe matters referred to in s 32(c)(i) and (ii). That is a question of fact.[43] Suppose a drug is administered which, as a result of an unanticipated allergicreaction causes injury. It is clear that there has been treatment injury. The injury wasplainly not the ordinary consequence of the treatment, which would never have beenadministered had the consequence been anticipated.[44] Next, take the case of a nerve cut during surgery, the nerve being in anunanticipated position. That will be covered because there could be no argument thatthe cutting of the nerve was an ordinary consequence of the surgical procedureundertaken.[45] In both these kinds of case the factual issue raised by s 32(1)(c) can easily beanswered. There is no doubt what has occurred is not a necessary part of the treatment.In other cases the issue may not be so easily determined.[46] Another kind of case may involve a course of treatment properly embarkedupon, but badly carried out. Taking as an example a surgical procedure that could becarried out in a way that does not cause a particular injury; the surgeon attempts theprocedure but nevertheless causes the injury. Whether there has been treatment injurydepends on a conclusion that the operation ordinarily could have been carried outwithout causing the injury. In other words, the operation was not performed as itshould have been. In this kind of case, establishing the injury was not a necessary partof the treatment involves showing that there was a better way of going about thetreatment. In any given case there might be a factual dispute about this, and resolvingthe dispute in favour of cover depends upon showing a proper standard was not met.So, establishing there was a treatment injury may involve showing that a professionalstandard has not been met.[47] A similar conclusion can be reached in the case of treatment injuries based onother kinds of treatment set out in subsequent paragraphs of s 33(1). Take the case ofa diagnosis of a person's medical condition, referred to in s 33(1)(b). For a diagnosisto cause personal injury it must, we think, be a wrong diagnosis. It would be a strangereading of the provision to suggest that it envisaged a correct diagnosis. That wouldmean that, applying s 32(1)(b), a correct diagnosis had caused personal injury. Itseems most unlikely that can have been intended. We consider it much more likelythat the intent was to embrace misdiagnosis. This too connotes a departure from astandard: a standard that would have been met had a correct diagnosis been given.[48] We consider next s 33(1)(c). This refers to a decision on the treatment to beprovided. Once again, we think the statute must have in contemplation an incorrectdecision as to the treatment to be provided. A correct decision about the treatment tobe provided cannot logically result in treatment injury because, if there is any injuryas a consequence of the decision, it will notionally be injury that is a necessary part orordinary consequence of the treatment. Were that not so, the wrong decision on thetreatment to be provided would have been made. In that case, any treatment injuryresulting from acting on the decision about the treatment to be provided would beinjury caused by the way in which the treatment was administered, and not the resultof the decision on the treatment itself. There would be a treatment injury arising fromthe giving of the wrong treatment.[49] Again, we think this provision turns on the correctness of the decision on thetreatment to be provided and that necessarily incorporates a standard. If the rightdecision on the treatment to be provided is made, that decision will not be the causeof any personal injury that results.[50] This brings us to s 33(1)(d): "a failure to provide treatment, or to providetreatment in a timely manner". It is clear that the word "failure" relates to both partsof this paragraph: the provision refers to both a failure to provide treatment, and afailure to provide treatment in a timely manner.[51] A point that can then be made is that the reference to provision of treatment ina "timely manner" of itself necessarily incorporates a standard. It involves a judgmentthat something was not done as soon as it should have been done. This must turn onsome standard that can be referred to in order to make the judgment. It is inherentlyunlikely in our view that "failure" in this context was intended to connote only thesimple idea that the appropriate treatment was not given earlier; to say in this contextthat something was not done in a timely manner involves a judgement that it shouldhave been done earlier.[52] If that is right, we consider a similar approach must be taken to "failure" in thefirst part of the paragraph. Here, the standard is not established by reference to timingbut simply by reference to a step not taken. In this setting, we consider a standardmust again be involved as a reference point. In order for there to be a failure to providetreatment, we consider there must have been some indication at the time of the failurethat the treatment not provided should have been provided. Unless that approach istaken it is difficult to see how there could be a relevant failure to provide treatmentthat caused the personal injury.[53] Mr Butler emphasised that the words "caused by treatment" in s 32(1)(b), readtogether with s 32(2)(a) (which excludes personal injury wholly or substantiallycaused by a person's underlying health condition) show that a failure to providetreatment under s 33(1)(d) on its own would be insufficient to attract cover. A failureto provide treatment must in fact cause injury. Here, Mr Butler relied on this Court'sdecision in Cumberland v Accident Compensation Corporation.39[54] In that case, this Court said that on a "traditional" or "standard" approach tocausation, in a case where it is alleged that a failure to diagnose has resulted in personalinjury, the relevant question to ask is whether "if the diagnosis had been properlymade, and proper treatment had followed, was the patient more likely than not to haverecovered".40 The Court proceeded on the basis that for a failure to be causative ofinjury, it was necessary to prove on the balance of probabilities that an alternativetreatment that would have prevented the injury would have "actually taken place".41Mr Butler argued that an alternative treatment must have been both clinically availableand preferable to the treatment in fact given before it could have been said that theinjury would have been prevented by an alternative course of treatment. The necessarycausal element depends on what is clinically indicated: a course of action that is notclinically indicated could not be said to be available.[55] We accept that argument. As a consequence, it would only be where there wasa departure from what was clinically indicated that a failure to provide treatment couldbe regarded as having caused injury in terms of s 32(1)(b).[56] Further, it is necessary to bring s 32(1)(c) into account. We agree withMr Butler that in order to ascertain whether a treatment injury was a necessary part orordinary consequence of treatment, it will often be necessary to consider clinicalindications relevant to the treatment. In our view, if treatment is not provided, in theabsence of clinical indications that there should have been treatment, para (c) meansthat there would be no treatment injury. This, because the injury that followed thenon-treatment would be a necessary part or ordinary consequence of the non-treatment39 Cumberland v Accident Compensation Corporation [2013] NZCA 590, [2014] 2 NZLR 373.40 At [50].41 At [45]–[46].taking into account the person's underlying health condition and the clinicalknowledge at the time. The reference to clinical knowledge at the time is telling. Interms of s 32(1)(c)(ii) not providing the treatment would have been in accordance withthe clinical knowledge at the time.42[57] Further, as Mr Butler submitted, an assessment of what was an ordinaryconsequence of treatment must involve reference to the expected outcome of treatmentgiven in accordance with proper medical practice. The fact that s 32(1)(c) refers tothe underlying health condition of and clinical knowledge about the patient as mattersto be taken into account to judge whether the injury was not a necessary part orordinary consequence of treatment supports a similar approach to "failure" and"timely" in s 33(1)(d).[58] It is also the case that non-treatment results in the development of a person'sunderlying health condition, which is excluded from the definition of treatment injuryby s 32(2)(a). In the absence of clinical indications for treatment of the underlyingcondition we do not think it can be said that the resultant injury is a necessary part orordinary consequence of the non-treatment taking into account the person's underlyinghealth condition at the time.[59] It is not necessary for present purposes to discuss every paragraph in s 33(1),but we mention s 33(1)(g) because Mr de Groot relied on it as another example of thelegislature providing for failure simpliciter, without reference to a standard. Unlikethe other paragraphs in the subsection, the focus of para (g) is on the failure ofequipment, and not the actions of health professionals. But we do not think it accurateto regard it as providing for failure simpliciter. Machines are designed to fulfilparticular purposes. If they malfunction they have failed to do what they weredesigned and intended to do. What they are designed and intended to do is the standardnot met and failure is assessed in terms of that standard. This is another example ofthe standard simply being implicit in the legislative idea being expressed.42 We note that in McEnteer v Accident Compensation Corporation [2010] NZCA 126, [2010]NZAR 301 at [20] this Court held that the question whether an injury that occurred during surgerywas a necessary part or ordinary consequence of the treatment was to be answered on the basis ofexpert opinion given that reflected on what had actually occurred, not in the abstract.[60] The foregoing analysis, focused on the text of the relevant statutory provisions,leads us to reject Mr de Groot's arguments. We are not persuaded that theinterpretation of the particular statutory provisions relevant to this case should beinfluenced by the provisions elsewhere in the Act on which he relies as examples ofwhere the legislature had included an explicit reference to a relevant standard. In thepresent case, a departure from what should have been done is inherent in thecombination of ss 32 and 33 for the reasons we have given.[61] In the result, we consider the relevant provisions properly construed mean thatin order for there to be treatment injury as a result of a failure to provide treatment itis necessary to show that an alternative treatment that would have prevented the injurysuffered could and should have been given having regard to the clinical indications atthe time of the alleged failure.[62] Taken as a whole the provisions indicate a legislative intent to limit cover forpersons who suffer injury while undergoing treatment, rather than providing cover forall those who suffer. The injury said to be a treatment injury must be the consequenceof a departure from appropriate treatment choices and treatment actions. The draftingcould have simply provided for cover for all injury suffered while a person undergoestreatment. But that course was not taken. Rather, boundaries were set out that havethe effect of limiting the availability of cover for injury during treatment. A failure inthe sense of omitting to take a step required by an objective standard is necessary.[63] We are not persuaded there is necessarily any inconsistency between theseconclusions and the various policy considerations on which Mr de Groot relied. Forexample, we do not consider the Hansard extracts to which he referred justify theclaims made about the intent of the legislature in passing the 2005 amendment. It mayaccurately be said, as the Minister claimed, that a consequence of the amendment wasthat ACC was no longer obliged to make findings of fault, but that is not the samething as saying that ACC does not have to be satisfied that an alternative course wasavailable that should have been pursued. Similarly, with respect to the additionalmaterials on which reliance was placed in this Court, while we accept that they indicatean intention to remove the previously applicable hindsight exclusion we do not acceptthat is conclusive for the interpretative exercise we are required to undertake.[64] In any event, Mr Butler properly relied on the content of the explanatory notewhen the 2005 amendment was introduced. The conclusions we have reached are notinconsistent with the relevant policy aim of moving "away from the punitive systemof finding medical error and reporting medical misadventure cover decisions".43 Andwe consider Mr Butler rightly drew a distinction between the attribution of blame toindividual professionals and determining whether treatment accorded with what wasclinically indicated in given circumstances.[65] As is always the case, it is necessary to focus on the words Parliament hasactually used. It will be apparent from our reasoning that we have discerned alegislative policy that, while not requiring a finding of negligence, still operates on thebasis that a treatment injury will only have occurred where there has been somedeparture from a standard and that departure has caused a personal injury.[66] There is a clear difference in academic opinion on these issues. Our conclusionis consistent with some of the views expressed by Professor Todd in The Law of Tortsin New Zealand.44 Professor Manning reaches different conclusions, rejectingnegligence-based reasoning and embracing the application of hindsight to theassessment of whether there has been treatment injury.45[67] Professor Todd recognises that while the legislation has abandoned anyreference to "mishap" or "error" in the treatment injury provisions introduced in 2005,both concepts may in fact necessarily be involved in a decision as to whether atreatment injury has occurred.46 He illustrates this by reference to s 32(1)(c), notingthat it will require the court to decide whether, in a given case, unwanted side effectsof a treatment are necessary or ordinary. Although those questions now replace theprevious concept of rarity, the courts must once again "determine whether there hasbeen some kind of mishap justifying coverage".4743 Injury Prevention, Rehabilitation, and Compensation Amendment Bill (No 3) (165-1)(explanatory note) at 2.44 Stephen Todd "Accident Compensation and the Common Law" in Stephen Todd (ed) The Law ofTorts in New Zealand (7th ed, Thomson Reuters, Wellington, 2016) 21.45 Manning, above n 27.46 Todd, above n 44, at 54.47 At 54–55.[68] Professor Todd also refers to the requirement in s 32(2)(a) that the personalinjury should not be wholly or substantially caused by a claimant's underlying healthcondition. He writes:48Where there is a failure to treat and the patient's condition gets worse, ortreatment does not alleviate a condition, how do we determine whether thecontinuing injury is caused by the treatment or by the underlying condition?Seemingly in each case a claimant must establish on the balance ofprobabilities that treatment, or different treatment, would have improved thepatient's condition or prevented it from getting worse.[69] He goes on to suggest that the now replaced concepts of medical mishap andmedical error "implicitly remain highly relevant", and potentially "cover theground".49 We consider these observations go further than justified by the treatmentinjury provisions of the Act. A distinction can be drawn between finding that adifferent treatment course should have been followed and establishing there wasnegligence. While both conclusions would often happily co-exist, the negligencefinding is not one required by the Act. More importantly, we consider an approachfocused on negligence reasoning runs the risk of failing to meet the legislative purpose.In some circumstances it may be that a higher standard is appropriate than thatcaptured by the concept of breach of a standard of care if the reference point is simplyto skilled and informed practitioners generally operating in a field.[70] Despite her rejection of reasoning based on negligence, Professor Manningrecognises that the legislation notionally involves an appropriate standard againstwhich to assess the treatment that has been provided or withheld in the case ofparticular claimants. In her article referred to above, she says:50it is entirely logical that, in order to reach a conclusion on whether there hasbeen a failure or delay in diagnosis or treatment causative of injury, one needsto measure what actually occurred against an appropriate benchmark; and todetermine what might otherwise have been the course of the patient's injury,given appropriate treatment or care.[71] We agree. These observations tend to support our conclusion that the decisionas to whether there has been a treatment injury will often turn on whether some other48 At 55.49 At 55.50 Manning, above n 27, at 36.course of treatment should have been taken other than the treatment in fact providedor withheld. But on the Court's approach there must be some indication of the needto take the alternative course before the course taken can be regarded as causing theinjury.[72] Professor Manning goes on to discuss Scandinavian patient schemes underwhich compensation depends upon whether an injury could have been avoided orprevented. She notes, however, that under those schemes compensation is notcontingent on, and no assessment is made of, whether the injury was due to error ornegligence on the part of the medical professional. Rather:51The guiding principle used to determine whether or not an injury wasavoidable is "the experienced specialist standard". According to thisprinciple, compensation will be paid for an injury if it can be assumed that anexperienced specialist in the field in the circumstances would have acteddifferently during examination, assessment or treatment, at the time ofassessment of the claim, thereby avoiding the injury. The classic faultprinciple is not relevant.[73] There was no suggestion of a breach of an experienced specialist standard inthis case. There were no observable indications prior to the fever that warrantedmedical intervention and there were no indications for a different treatment course inthe circumstances. We are bound to act on those factual findings, set out in the casestated.Result[74] The application to adduce further evidence is granted.[75] We do not propose to answer all of the questions posed in the case stated. Wethink it is unnecessary to do so, and there are assumptions behind some of them thatin our view would make it unsafe to answer them in the form they have been posed.We think the better course to follow is simply to answer question (a) in the affirmative.The reasons for reaching that conclusion have effectively dealt with the variousmatters raised in argument, including the key underlying issue as to the properinterpretation of s 33(1)(d).51 At 37–38 (footnote omitted).[76] We answer question (a) of the case stated (was the High Court correct to findthat a failure to treat cannot occur in circumstances where there are no indications fora different treatment course?) yes.[77] The appeal is accordingly dismissed.[78] The issue of costs was not raised in written submissions or at the hearing. Ourprovisional view is that this being in the nature of a test case costs should lie wherethey fall. If despite that indication ACC wishes to seek costs, a memorandum is to befiled on that issue within 10 working days of the judgment. Any memorandum in replyfrom the appellant is to be filed five working days after receipt of ACC'smemorandum.Solicitors:John Miller Law, Wellington for AppellantRussell McVeagh, Wellington for Respondent