Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
Patients don't take the prescribed dose of inhaled steroids
- C J Cates (19 October 1998)
-
Pitfalls in analysis of inhaled steroid prescribing.
- Peter Black (26 October 1998)
-
Analysis has important shortcomings
- Staffan Edsbäcker (5 November 1998)
-
Methodological limitations in epidemilogic studies of asthma medication potencies
- Samy Suissa (24 November 1998)
-
Response to Dr. Cates Re: Patients don't take the prescribed dose of inhaled steroids
- B D Pethica (4 February 1999)
-
Response to Dr. Black Re: Pitfalls in analysis of inhaled steroid prescribing.
- B D Pethica (4 February 1999)
-
Response to Dr. Edsbacker Re: Analysis has important shortcomings
- B D Pethica (4 February 1999)
-
Response to Prof. Suissa Re: Methodological limitations
- B D Pethica (4 February 1999)
-
Further evidence of potency differences between beclomethasone and budesonide?
- Alister Penrose (20 August 1999)
|
|
|||
|
C J Cates, General Practitioner Manor View Practice, Bushey Health Centre, Bushey, Hertfordshire
Send response to journal:
|
EDITOR- The conclusions reached by Pethica et al are not justified because they have not verified an important assumption in their paper concerning the relative potency of inhaled beclomethasone and budesonide. They have assumed that patients are regularly taking the dose recorded on their prescription and the data presented in their paper suggests that this is most unlikely. The total number of prescriptions issued to 5930 patients was only 16,725 (less than three per year). Most formulations of inhaled steroid last 50 days at the usual prescribed dose, so patients taking regular treatment should have used around seven prescriptions in the twelve months of the study. Analysis of computerised prescriptions in our practice over the twelve months of July 1997 to June 1998 shows a similar pattern; 625 patients received 2424 prescriptions for inhaled steroids, (around four per year on average). Analysis of individual data on 59 patients aged 30 to 35 years shows that 36/59 were prescribed less than four inhalers per year. These patients did not receive enough medication to have taken the prescribed dosage of inhaled steroid regularly over the twelve-month period. The analysis carried out in Pethica’s paper is based on the mean prescribed dose and this may bear little relation to the dosage actually taken over the whole period. The total inhaled dose for each patient would be better estimated by multiplying the contents of each inhaler by the number of inhalers prescribed over the twelve months studied. Analysis should be restricted to those patients who can be shown to be taking regular inhaled steroids. Beclomethasone is available as a 100mcg metered dose inhaler (mdi), whilst budesonide is only available for adults as a 200mcg mdi. Since general practitioners are sometimes creatures of habit and a dosage of 2 puffs bd is commonly use when starting inhaled steroids, there is a possibility that budesonide in adults may be started at higher dosage. This could distort the findings in Pethica’s analysis. Before any guidelines are altered further analysis is required of Pethica’s data, as well as confirmatory evidence from patients in other countries. Prescriptions for inhaled steroids issued per patient in our practice from July 1997 to June 1998 (patients aged 30-35 years). Inhalers/yr 1 2 3 4 5 6 8 9 10 or more No of pts 20 12 4 5 2 7 2 1 6 |
|||
|
|
|||
|
Peter Black, Senior Lecturer in Medicine University of Auckland, New Zealand
Send response to journal:
|
On the basis of a retrospective analysis of prescriptions by general practitioners in New Zealand, Pethica et al suggest that budesonide is less potent than beclomethasone dipropionate. This is not an appropriate way to assess the relative potencies of inhaled steroids and the conclusions are potentially misleading. One could draw this conclusion if the patients were randomised to treatment with budesonide or beclomethasone and the dose was carefully titrated so that the minimum effective dose was used. In clinical practice patients are not randomised to treatment and,in my experience, back titration of inhaled steroids is the exception rather than the rule in New Zealand. A controlled trial which did compare budesonide from the turbuhaler with beclomethasone dipropionate from a metered dose inhaler (MDI) and spacer found that 600 mg of budesonide had a similar effect to 1000mg of beclomethasone (Allergy. 1994;49:833-36). It is unlikely that the high doses of budesonide are used in New Zealand because budesonide is less potent than beclomethasone. There are, however, plausible explanations for the findings reported by Pethica et al. I have observed that patients in New Zealand are often switched from the beclomethasone MDI to the budesonide turbuhaler when their asthma is poorly controlled. This is likely to lead to the prescription of higher doses of budesonide than beclomethasone. The use of higher doses of budesonide has also been encouraged by the availability of the 400mg/dose turbuhaler whereas the highest dose available from the beclomethasone MDI is 250mg per actuation. (Although beclomethasone is available as 400mg per actuation from the diskhaler this is prescribed infrequently in New Zealand.). When a patient is commenced on high dose inhaled steroids it is clearly much more convenient to prescribe 2 puffs twice a day of budesonide from a turbuhaler than it is to prescribe 4 puffs twice a day of beclomethasone from an MDI. Pethica et al argue that their findings are not confounded by severity because the subjects on budesonide were prescribed the same number of courses of prednisone as those on beclomethasone. One could equally well argue that the reason the subjects on budesonide did not have more exacerbations was because they were prescribed a higher dose of inhaled steroid. To exclude confounding by severity they should have been assessed the severity of the asthma prior to treatment with inhaled steroids. Pethica and colleagues particularly stress the use of higher doses of budesonide in all age groups. The differences in the average doses of the budesonide and beclomethasone in children is easily explained when one appreciates that the lowest dose form of budesonide is 100mg per actuation compared with 50mg per actuation for beclomethasone. The recommendation that budesonide should be prescribed in higher doses than beclomethasone is likely to be incorrect. It is also unhelpful but it could dissuade doctors from back titration in patients who are on treatment with high doses of inhaled budesonide and who have well controlled asthma. |
|||
|
|
|||
|
Staffan Edsbäcker
Send response to journal:
|
Editor - The suggestion by Pethica et al. (1) that budesonide is less potent than beclomethasone dipropionate (BDP) is not valid. Prescription pattern studies may, at best, generate hypotheses. However, the following important shortcomings in the study by Pethica et al. exclude even this possibility: The authors have not shown the 28 New Zealand centres to be representative for the country. Centre effects and differences in prescription patterns are not described, e.g. the doses were much higher (approximately 1000 g) than those currently recommended (2). Drug compliance and actual drug exposure have not been validated. The patients were not matched for severity of asthma before the treatment was started. Pethica et al claim that the data does not support confounding by severity, yet a significantly higher proportion of patients receiving budesonide accessed hospital services (7% vs 4%, p=0.008). Also, the proportion of budesonide patients who had received secondary care (56%) was markedly higher than that of the entire patient material (35%). No clinical outcome data are provided. Pethica et al. claim that back titration was common practice among the doctors but present no data, despite access to computerised records. As systemic effects are more marked with BDP (1), the inclination to follow a back titration scheme should be greater with this drug. Pethica et al. suggest that the lower receptor affinity of budesonide vs BDP supports a lower systemic effect in addition to a postulated lower therapeutic effect. Obviously, any prediction of therapeutic or adverse effects based on laboratory tests of topical anti-inflammatory potency is highly speculative, since such tests do not account for either physicochemical or pharmacokinetic properties of individual drugs nor delivery system performance. It is well established that Turbuhaler® inhaler performs better than pMDIs (3) and that the oral availability is less for budesonide (10%) than for BDP (around 30%) (4). In addition, in comparing topical potency, the skin blanching assay (ratio BUD/BDP 1.63) is at least as relevant as in vitro receptor binding affinity (ratio BUD/BDP 0.70) (2). Therapeutic equipotency of budesonide and BDP via pMDI has been documented in well controlled clinical studies. Budesonide delivered via Turbuhaler® inhaler is twice as potent (3,4). The study by Pethica et al. (1) can be characterised as a description of prescription habits in a selected set of centres. It cannot be used to assess relative potency; this requires prospective, randomised, controlled clinical trials. Staffan Edsbäcker, PhD, Associate Professor Experimental Clinical Pharmacology University of Lund, Lund, Sweden Associate Director, Human Pharmacology, Clinical Research & Development, Astra Draco AB S-221 00 Lund, Sweden Maria Gerhardsson deVerdier, MD, Phd, Associate Professor of Epidemiology Karolinska Institute, Stockholm, Sweden Epidemiologist, Clinical Drug Safety & Epidemiology Clinical Research & Development, Astra Draco AB Lund, Sweden Christer Hultquist, MD, Medical Advisor Clinical Research & Development, Astra Draco AB Lund, Sweden 1 Pethica B D, Penrose A, MacKenzie D, Hall J, Beasly R, Tilyard M. Comparison of potency of inhaled beclomethasone and budesonide in New Zealand: retrospective study of computerised general practice records. BMJ 1998; 317: 986-990. 2 Murphy S, Bleecker ER, Boushey H, Buist AS, Busse W, Clark NM et al. Guidelines for the diagnosis and management of asthma. Expert panel report II 1997. 3 Barnes PJ, Pedersen S, Busse WW. Efficacy and safety of inhaled corticosteroids. New developments. Am J Respir Crit Care Med 1998: 157 (No 3 Pt 2): S1-S53 4 Boobis AR. Comparative physicochemical and pharmacokinetic profiles of inhaled beclomethasone dipropionate and budesonide. Respir Med 1998; 92 (Suppl B): 2-6 5 Makino S, Miyamoto T, Takahashi T, Nakajima S, Yamakido M, Mano K et al. Multicentre study comparing budesonide Turbuhaler with beclomethasone dipropionate (BDP) pMDI in Japanese patients with asthma. Chest 1996: 110 (Suppl No 4):84S |
|||
|
|
|||
|
Samy Suissa, Professor of Epidemiology and Medicine McGill University Health Centre
Send response to journal:
|
To the Editor: The study by Pethica et al, by its cross-sectional design, could have equally arrived at a different conclusion. Indeed, the authors could very well have concluded from these data that higher doses of inhaled steroids given to more severe asthmatics lead to a reduction of oral steroid use and outpatient consultations to a level equal to that of moderate asthmatics. Since data on inhaled steroids were obtained during the same one-year period as the data on asthma severity markers (namely oral steroid use, outpatient consultations and emergency room visits) it is impossible to determine which came first. It is therefore equally possible that these markers of severity were altered by the high dose inhaled steroids. An important methodological limitation inherent in the study design may have also biased the findings. For metered-dose inhalers, for example, beclomethasone was available in canisters of 50, 100 and 250 mcg per puff, while for budesonide it was only available with 200 mcg per puff. By this mere fact, the mean dose of inhaled steroids will be shifted downwards by the lower available doses for beclomethasone. Moreover, because the lower doses for beclomethasone (50 mcg) are more likely to be prescribed for flexible regimens and the authors decided to use the minimum values for these types of prescriptions, the mean dose of beclomethasone will be shifted down further. Thus, the nature of the available formulations imposed a bias in the comparison, even before the data were collected. Another indication of bias stems from the number of prescriptions per subject per year which is roughly equal for beclomethasone (12,305/4,925=2.5 per year) and budesonide (3,957/1,532=2.6 per year). Since the average daily dose is 50% higher for budesonide (979 vs 635 mcg per day), the duration of use of each canister must be shorter for budesonide than for beclomethasone. Specifically, the daily dose of 979 for budesonide implies that less than 5 puffs were prescribed per day. The average of 2.6 canisters per year implies that 520 puffs were available, so that the subjects on budesonide used their medication for around 105 days per year. These calculations are not possible for beclomethasone because of the different inhalers, but the rates indicate that the average duration for beclomethasone users may be longer than for budesonide users, which introduces another source of bias in the comparison between the two. Some data from this study are disturbing. For example, 11 patients aged 0-2 years received 30 prescriptions of budesonide with a mean daily dose of 582 mcg. Using the standard deviation of 260 and a conservative assumption of a bell shape distribution for the dose, we note that a child used around 1,100 mcg of budesonide per day. Did that child use 1,100 mcg per day for 105 days? This appears rather excessive for that age and should perhaps be investigated. In all, this study has several methodological limitations, the most important being that the design is cross-sectional so that a contrary conclusion, indicating the effectiveness of higher doses of inhaled steroids in reducing asthma outcomes, is just as tenable from these data. Samy Suissa Professor of Epidemiology & Biostatistics and of Medicine, McGill University Director, Division of Clinical Epidemiology, Royal Victoria Hospital Pierre Ernst Professor of Epidemiology & Biostatistics and of Medicine, McGill University Medical Scientist, Division of Clinical Epidemiology, Royal Victoria Hospital Montreal, Canada NOTE: The authors' asthma research is funded by Astra, Boehringer Ingelheim and Glaxo-Wellcome. |
|||
|
|
|||
|
B D Pethica, research fellow Wellington Asthma Research Group
Send response to journal:
|
Dear Dr. Cates, Thank you for your rapid response commenting on the conclusions in our "Comparison of potency of inhaled beclomethasone and budesonide in New Zealand...."BMJ 1998; 317: 986-90. We did not, as you suggest, make any assumptions about compliance or that therapy was taken regularly or continuously. Our key assumption is clearly stated, namely that practitioners adjust the dose of inhaled corticosteroids according to clinical response. This is standard practice in New Zealand. At the time of our study it was usual to dispense up to 3 months of therapy, and your suggestion that seven prescriptions per annum would be needed for continuous therapy is not applicable to NZ. While compliance is plainly an important determinant of clinical response, ascertaining actual compliance is a problem for all efficacy studies of inhaled corticosteroids in man. Differential compliance can importantly bias a randomised controlled trial to compare potency. The difference in mean prescribed dosage which we found in our study would require substantially worse compliance with the budesonide formulations than with the beclomethasone formulations. There is no literature to our knowledge which would convincingly support such a differential compliance. More importantly, our analyses showed that the difference in prescribed daily dosage was consistent for different device types. If compliance was an important factor there should be variation of the difference in prescribed daily dose according to the device type, which was not the case. Therefore differential compliance is very unlikely to explain the dose difference we report. Interpretation of the time interval between prescriptions is complicated by the fact that not all patients are taking continuous inhaled corticosteroids in the calendar year analysed. Some patients start and others stop, as you describe from your own practice, with many appearing to use inhaled corticosteroids continuously. The time interval also depends on the prescribed quantity e.g. most MDIs and the Turbuhaler have 200 doses and Diskhaler 120 doses, but multiples can be dispensed, and when the prescribed single doses do not tally with the device single dose (e.g. if 800mcg bd or more is prescribed), this makes interpretation of the data particularly awkward. This is one reason why our focus in the manuscript has been on the mean daily prescribed dose; another reason is that this best takes into account the patient’s clinical state at the time of prescription. Overall, in our data there does not appear to be evidence of an important difference in the time interval between prescriptions for beclomethasone and budesonide. While it may be superficially attractive to limit the analyses to patients who can be shown to be taking their inhaled corticosteroids regularly, it is currently impossible to say which patients these are. Such analyses are fortunately of little importance when the underlying time intervals between prescriptions for budesonide and beclomethasone are similar. It is also noteworthy that, in randomised trials, limiting the analysis to treatment compliant patients can very easily lead to false conclusions, which is why intent to treat analyses have become standard under most circumstances rather than observed case analyses. The available formulations and dosages in NZ are listed in our manuscript. You suggest that low initial doses because of MDI formulation dose differences could importantly affect the population mean prescribed daily dose. The magnitude of the mean daily doses which we found makes this unlikely, and it would not explain the consistency of the dose difference for the different device types. Two puffs twice daily can be used to deliver quite a range of daily doses for the commonly prescribed Diskhaler and Turbuhaler breath activated devices, with the same nominal product dosing available for both. The mean doses of budesonide were higher than those for beclomethasone for both MDIs and breath activated devices. We stand by our conclusions based on our data, but agree, as stated in the conclusions section of our manuscript, that confirmatory evidence is now needed. |
|||
|
|
|||
|
B D Pethica, research fellow Wellington Asthma Research Group
Send response to journal:
|
Dear Dr. Black, Thank you for your rapid response letter regarding "Comparison of potency of inhaled beclomethasone and budesonide in New Zealand...."BMJ 1998; 317: 986-90. There are numerous difficulties in achieving a satisfactory design for clinical trials to compare the potency of inhaled steroids for the treatment of asthma. Some of these problems have been outlined in the introduction section of our manuscript. ‘No difference’ results from randomised trials are hard to credibly interpret in the absence of a negative control group to prove sensitivity. Insensitivity may give ‘no difference’ results across a wide dosage range. The rate of switching between budesonide and beclomethasone was low in our study (103 patients switching from one to the other). This number is too small to have a substantial effect on the mean prescribed daily dose. The number of patients prescribed Becodisk in our database was 1100, and Pulmicort Turbuhaler 1414. The Diskhaler was almost as commonly prescribed as the Turbuhaler, and both devices have 100mcg, 200mcg and 400mcg per dose formulations, giving no difference in convenience. The 50 mcg per actuation formulation of beclomethasone was mostly restricted to age 6 years and under in our data, which we accept may introduce a bias in this age group, accounting for the lower relative rate of budesonide prescribing in this age group. However, exclusion of the 50mcg formulation from the overall analysis has minimal effect on the mean prescribed daily dose. The relative merits and demerits of using asthma severity markers from a previous year for our analysed patients would make an interesting debate given the inherent variability of asthma. Secondary care data from the previous year (1 July 1993 – 30 June 1994) were analysed for the 1409 patients for whom linkage to secondary care data was possible. The results were similar to those as presented in our publication (Table 3): patients treated with budesonide were more likely to have any secondary care events, but there was no significant difference in secondary care event rates amongst patients who had received two or more high dose prescriptions. Our conclusions would therefore be unchanged by such an analytic approach. However, access to secondary care data for the whole of the studied population would plainly be preferable. Hopefully this will be possible as part of the confirmation process which we consider to be required. |
|||
|
|
|||
|
B D Pethica, research fellow Wellington Asthma Research Group
Send response to journal:
|
Dear Dr. Edsbacker and colleagues, Thank you for your www.bmj.com rapid response letter regarding "Comparison of potency of inhaled beclomethasone and budesonide in New Zealand...."BMJ 1998; 317: 986-90. We do not agree that prescribing pattern studies can only generate hypotheses. You describe inhaled corticosteroid doses above 1000mcg per day as "much higher" than those currently recommended. A prescribing pattern study can definitively establish that such prescriptions are far from uncommon. Indeed, prescribing at these doses is recommended in the New Zealand product information for budesonide Astra Draco. Observational studies such as ours can usefully address numerous questions very difficult to answer using randomised clinical trials; some of the difficulties in the design of randomised trials to credibly compare the potency of inhaled corticosteroids in asthma are covered in our manuscript. Randomised clinical trials comparing different doses of different inhaled corticosteroids have generally resulted in ‘no difference’, making them hard to meaningfully interpret. Our alternative approach has been to study what is actually prescribed in New Zealand General Practice, and seek explanation for the higher mean prescribed doses of budesonide compared to beclomethasone. Our manuscript covers some of the potential sources of confounding, specifically in relation to the representativeness of the practices, and our interpretation of the hospital service access difference to which you refer. We do not claim that back titration was necessarily implemented, but dose adjustment. The interpretation of the data would be logically similar if no back titration actually occurred, but only dose increase in response to poor asthma control. Nevertheless, back titration has been promoted widely in New Zealand and is standard practice. More important for our interpretation is that the state of a patient’s asthma when they are seen is reflected in the dose prescribed. We do not have compliance information, but refer you to our responses to Dr. Cates and Prof. Suissa on this topic. |
|||
|
|
|||
|
B D Pethica, research fellow Wellington Asthma Research Group
Send response to journal:
|
Dear Profs. Suissa and Ernst, Thank you for your www.bmj.com rapid response letter regarding "Comparison of potency of inhaled beclomethasone and budesonide in New Zealand...."BMJ 1998; 317: 986-90. The alternative explanation you propose for the higher mean prescribed doses of budesonide initially appears plausible – that budesonide is prescribed in higher doses to more severe asthmatics whose use of hospital services is therefore relatively decreased. This suggests that even though they are more severe asthmatics their severity markers should be decreased. However, as discussed in our manuscript, on restriction of the analysis to patients who received at least two prescriptions for higher doses of either budesonide or beclomethasone it appears that secondary care utilisation is similar. Increased secondary care utilisation occurred amongst budesonide patients prescribed lower rather than these higher doses. Combined with the post-study survey of the contributing general practices, all of whom believed that budesonide and beclomethasone were equipotent, this suggests lower potency of budesonide rather than systematic prescribing of budesonide to more severe asthmatics. Nevertheless, we accept that an observational study like ours has weaknesses which do not permit over-definitive conclusions. Clinical trials so far performed to compare the potency of these products mostly have grave methodological weaknesses. Confirmation of our conclusions, or their refutation, is needed from other studies, and an epidemiological approach would appear entirely appropriate. One aspect of such future studies could be to use the recognised asthma severity markers from a previous year (please see our reply to Dr. Black on this topic). Analysis of the secondary care data for 100% of the included population, which we were unfortunately unable to access, would also be much preferable. Regarding the MDI formulations, if the dose differences we found were limited only to MDIs we would fully accept that this was likely the result of bias as you describe. However, the Becodisk and Turbuhaler devices have formulations with identical single doses (100, 200 and 400mcg per dose), but the difference in mean prescribed daily dose in analysis restricted to these devices was similar to that found with MDIs, (Becodisk 575mcg, Turbuhaler 989mcg). Your calculations of puffs per day and duration of use are hard to follow. There are different strengths per dose for both beclomethasone and budesonide formulations, so that a higher daily dose does not imply shorter duration of use. A puff of budesonide might vary from 100mcg to 400mcg, for example, and a 200 dose device could last from 50 to 200 days when the daily dose was 400mcg. In addition, prescribing was not limited to one device or number of doses. The actual usage and compliance with the prescribed product is not possible to address satisfactorily using computerised prescribing records, but our data does not suggest an important difference in potential duration of usage of the two medications. In the absence of data on actual usage, we have interpreted the prescribed daily dose to be a reflection of the prescriber’s assessment of the state of the patient’s asthma at around the time of the prescription. The underlying distribution of doses is not Gaussian. The maximal prescribed daily dose of inhaled budesonide to any child aged 0-2 years was 800mcg. In addition, in alternative analyses in which the maximal prescribed dose was used (e.g. 4 puffs if the prescribed dose was 2-4 puffs), the difference in prescribed daily dose between budesonide and beclomethasone is maintained. If there is undetected severity bias in spite of our best efforts to detect it, or other biases exist which we have not considered, we hope that this will be reflected in the results of the confirmatory studies which we consider now to be required. Pending such studies we stand by the conclusions as stated in our manuscript. Authors of BMJ 1998; 317: 986-90 |
|||
|
|
|||
|
Alister Penrose, Junior Research Fellow RNZCGP Research Unit, Department of General Practice, University of Otago, Dunedin, New Zealand
Send response to journal:
|
Dear Sir A recent case study by the National Preferred Medicines Centre (Premec)showed some interesting results concerning the potency of budesonide and beclomethasone (New Zealand Family Physician 1999;26:21- 24). The case study was sent to all 2,844 practising general practitioners in New Zealand. The case presented was: "A 21 year old woman comes to see you because her asthma has been worse over the past 12 months. She is waking one to two nights a week because of cough and chest tightness and she is using her salbutamol inhaler four or five times a day. Her PEFR measured in the surgery is 390l/min (predicted 450l/min). The only other medicine which she is taking is the combined oral contraceptive." A total of 1,005 general practitioners responded. When asked what additional medicines they would prescribe, 97% indicated they would prescribe an inhaled corticosteroid, either alone or in combination. Further, 49% selected beclomethasone, 36% fluticasone, and 12% budesonide. Of most interest was that the mean daily prescribed dose suggested was 518mcg for beclomethasone, and 808mcg for budesonide. These results are remarkably close to that found in our paper (Pethica BD, Penrose A, MacKenzie D, Hall J, Beasley R, Tilyard M. Comparison of potency of inhaled beclomethasone and budesonide in New Zealand: retrospective study of computerised general practice records. BMJ 1998; 317: 986-990). Dr Peter Black, who also responded to the BMJ on our paper, commented on the Premec study, and I quote verbatim. "Similarly, one would anticipate the dose of budesonide would be similar to that of beclomethasone, not 56% higher. Ths case study mirrors clinical practice. Ths doses of fluticasone and budesonide are higher than those for beclomethasone. The response to this study suggests the higher doses of fluticasone and budesonide prescribed in New Zealand are not just due to the preferential prescribing of fluticasone and budesonide to patients with more severe asthma." One could conclude that the respondents to this case study assessed the patient's severity prior to making their prescribing decision, which he notes we could not allow for in our paper. Yet the results of this study reflect our own. Alister Penrose Junior Research Fellow RNZCGP Research Unit Department of General Practice University of Otago Dunedin New Zealand |
|||