Product:Sensura One-Piece Closed Shallow Convex Pouch

PUBLIC SUMMARY DOCUMENT

Product:SenSura One-Piece Closed Shallow Convex Pouch

Applicant:Coloplast Pty Ltd

Date of SPAP Meeting:29-30 July 2013

1. Proposed Listing on the Stoma Appliance Scheme

The applicant, Coloplast, sought the addition of a price premium to a product currently listed in Subgroup 1(c) of the SAS Schedule, the SenSura One-Piece Closed Shallow Convex Pouch (SAS code 9851L). The applicant proposed a unit price inclusive of a price premium over the benchmark unit price for Subgroup 1(c) ($4.433).

The price premium requested was for:

1. ‘The ability of the SenSura double layer baseplate for improving skin condition and use of accessories.
2. Abilities of the SenSura appliance to improve aspects related to quality of life, valued by utilisation of willingness-to-pay.’

1. Comparator

The applicant nominated a Hollister product listed in Subgroup 1(c) of the SAS Schedule, the ModermaFlex One-Piece Closed Shallow Convex Pouch (SAS code 9943H), as the comparator. This product is currently listed at a unit price of $4.707, inclusive of a price premium of $0.274 for the AF300 filter.

1. Background

An application requestinga unit price premium (higher than the unit price premium applied to this product from 1 July 2011) was considered by the Stoma Product Assessment Panel (SPAP) in November 2012 as part of the Group 1 Price Premium Review (SAS 2012-13 Budget measure). At that time the SPAP recommended the removal of the existing price premium on the basis of insufficient evidence for its justification – a recommendation that received Cabinet approval in February 2013 and came into effect on 1April 2013.

1. Clinical Place for the Product

The product is a one-piece closed appliance with a convex baseplate suitable for use by people with a colostomy.

1. SPAP Comment

Clinical Analysis

The Dialogue Study (DK175OS)

The Dialogue Study was considered separately by the Panel as it was pivotal to the economic analyses presented by the applicant in all six Coloplast SenSura submissions considered at this meeting. It was a non-comparative, single arm, 6-8 weekmulti-centre trial conducted across 21 countries in which 3,017 subjects were switched from their current device to an equivalent SenSura product and outcomes before and after were compared.

In the view of the Panel, the following flaws in study designseriously compromised the validity of the evidence provided by the Dialogue Study:

1. it was non-comparative and single arm in nature;
2. at the first (inclusion) visit in week one, patients were instructed on the study and the use of SenSura appliances. If it was stated at this visit that the SenSura range of products was likely to offer an improvement, then this could have influenced patient outcomes. Furthermore and in relation to the instruction given on the use of the new appliances, 15% of participating patients had never visited a stoma clinic and 53% had only visited when needed. Thus over two thirds of subjects had not had regular clinic visits and discussion by the nurse alone may have significantly improved outcomes. This may have related to a range of activities unrelated to the nature of the device – for example removal, cleansing etc. The extent of the nursing effect was not able to be determined by the Panel due to the non-comparative nature of the trial; and
3. there was no evidence providedregarding whether or not advice regarding the management of any peristomal skin conditions (PSCs) present at entry was given at visit one. As 60% of patients were deemed by the nurses to have a skin condition at entry and skin disorders did not constitute an exclusion criterion, it seems ethically indefensible that these patients were not provided with advice and/or treatment options – particularly for those patients with a DET score of 4 or more (moderate to severe disease). It was surprising that patients with severe skin conditions (250 or 8% who had an initial DET score of 7 or more) were not excluded from the trial. The Panel considered the question of the ethics of non-intervention in the presence of peristomal skin problems at entry to be critical, and the absence of any information in this regard placed great uncertainty over any impact of the new devices on skin improvement over the study period.

The Panel noted the statistically significant change in the Quality of Life (QoL) metric developed by Pietro (2005) – a mean change from 58.1 to 59.9 (standard deviation 6.9) obtained from the data of 2,672 of the 3017 patients recorded – but was of the view that a difference of 1.8 on a 100-point scale was unlikely to be clinically significant (ie. translate into any meaningful clinical difference in quality of life for patients using SenSura appliances versus prior use products).

The Panel then considered the results based on the peristomal skin assessment tool developed by Coloplast (the DET score) from which the applicant extends an economic analysis. The DET score measures both extent and severity in the three domains of discolouration, erosion and tissue overgrowth, with a maximum score of 15 corresponding to maximum severity. The following issues with DET scoring in the Dialogue Study were identified by the Panel:

1. The Coloplast Ostomy Skin Tool Handbook gives useful tips to minimise skin irritation etc. and this was presumably part of the advice given to patients at visit one;
2. There was wide inter-countryvariation in DET score changes, likelyas a result of inter-assessor variability (Jemec et al. 2011). The Panel noted that as was appropriate, initial and final DET scores for individual patients were measured by the same nurse at both visits;
3. As outlined above in relation to study design, the improvement in DET score is likely to have been influenced by the nature of the nursing intervention in the context of a clinical trial. The applicant attempted to disaggregate the nursing effect from the performance of the appliance by comparing the entry and completion DET scores (and corresponding degree of reduction) of the groups who had never, when needed and regularly visited a stoma clinic, and noting that there was no statistically significant difference between the three groups, thus concluding that nursing effect was not a factor. This method was unconvincing as it relied on the dual assumptions that a) the use of the pre-study appliance was optimised and b) the quality of nursing advice was constant; and
4. As previously stated, it would have been deemed unethical not to intervene at visit one for those patients with a moderate to severe PSC at entry.

The Panel noted that the applicant’s own submission conveyed that the impact of nursing cannot be differentiated from that of the appliance: ‘the Dialogue Study showed that evidence-based nursing and the use of an appropriate SenSura appliance reduced leakage [leading to PSCs] significantly compared to the pre-study appliance’. Data on file by Aaes states that ‘the possibility of separating the effect of evidence based nursing given at study start and SenSura performance has not been addressed.’

For these reasons, the Panel concluded that the use of the DET score as the primary input into the applicant’s economic model was highly problematic.

Finally, the Panel noted that there was no concrete confirmation provided in either the study report or any of the supplements that any one SenSura product (of the six forming the subjects of the applicant’s current round of submissions) formed part of the Dialogue Study. That any one product was included was instead a matter of assumption. It was the opinion of the Panel that a subgroup analysis, with subgroups defined on the basis of the product used, should have been possible given the large number of overall study participants.

Other Evidence Presented

The Panel noted that the applicant presented numerous other studies in support of its application, but that only two among them presented data pertaining to the performance of a SenSura one-piece closed system versus an appropriate comparator. These studies were DK109OS (comparator product Dansac Nova 1; test product known to be flat) and DK145OS (comparator product Hollister ModermaFlex; test product presumed to be flat as at the time of the study there were no convex products in the comparator range).

Study DK109OS, a two-week (one week per phase), open label, cross-over, randomised controlled trial involving 68 subjects, had a primary endpoint of patient preference. Secondary endpoints included security, ballooning, flexibility, discretion and skin condition. Several of the results were favourable to the SenSura product – for example 85% patient preference, a higher level of security (94% versus 76%, statistical significance uncertain), a lower rate of ballooning (7% versus 27% – though as the study dates from 2007, it was noted that the comparator product would not have contained the AF300 filter) and better ratings on tack, adhesion, flexibility and discretion. No significant difference in skin condition was detected.

Study DK145OS, a two-week (one week per phase) open label, cross-over, randomised controlled trial involving 69 subjects, alsotook patient preference as its primary endpoint. Secondary endpoints included security, ballooning, discretion and skin condition. Again, several of the results were favourable to the SenSura product, including 78% patient preference, a higher level of security and a lower rate of ballooning (6% versus 23% – again it was noted that the comparator product would not have contained the AF300 filter). No significant difference in skin condition was detected.

The data from these two studies, which it was noted cannot be verified because no study reports are available, suggest that the product is at least no worse than its comparator, and may possibly be better. The open and short-term nature of the trials and the subjective and variable nature of the patient preference outcome measure (which would never be given primacy over the objective measurements which would verify it)constitute limitations, but they are studies relevant to the submission and it is detrimental to the applicant’s case that they are not developed further.

The Panel noted that two pieces of data were presented by the applicant with specific regard to the performance of the new SenSura filter contained within the product under consideration. The first was a 2010 study concerned with ballooning frequency and time to ballooning. The test product in this study contained the previous SenSura ‘Wave’ filter and thus its relevance to the comparison with the comparator product, the Hollister ModermaFlex, is uncertain.

The second piece of data was provided by the Coloplast Laboratory Test, an in-vitro test measuring air flow. Results were as follows:

• air flow through the Hollister AF300 filter was almost double that of the SenSura filters (both old and new designs) at the start;
• following one artificial contamination cycle the air flow through the test and both comparator products (the Hollister AF300 and the old SenSura wave filter) was the same; and
• following subsequent contamination cycles the new SenSura filter outperformed both comparator products.

These results were of interest to the Panel, however in the absence of an in-vitro/in-vivo correlation or a head-to-head study of the in-vivo performance ofa product containing the new SenSura filter versus a comparator containing the AF300 filter, interpretation of the results was uncertain.

Economic Analysis

The Panel noted that the economic case presented by the applicant in support of the price premium claimedhad three main components: a base case model based on the treatment costs associated withPSCs, a reduction in accessory use, and ‘willingness to pay’.

Peristomal Skin Conditions

The PSC arm of the applicant’s economiccaseused a 15 health state-classificationsystem based on combining five disease types (irritant-contact dermatitis, allergic dermatitis, mechanical trauma, disease-related and infection-related) with three degrees of severity (mild, moderate or severe – classified according to DET score). The differencein the proportion of patients with each type and severity of disease between visits one and two in the Dialogue Study was multiplied by the estimated cost of managing that specific combination of type and severity, then summed for all states and divided by the estimated product usage during the trial period (based on maximum quantitythresholds under the SAS) to calculate the PSC-based component of the per unit price premium.

As outlined above (cf. Clinical Analysis), the use of changes in DET score in the Dialogue Study to underpin this modelling is highly problematic due to likely considerable confounding – ie. stated percentage changes in DET score (the largest in the area of moderate/severe PSCs) cannot be interpreted as being solely due to the performance of the SenSura product. In the view of the Panel, the following additional flaws were inherent in this approach:

1. The incidence figures were obtained from the applicant’s financial worksheet and could not be verified in the absence of the raw data;
2. The costs associated with the treatment of mild, moderate and severe disease were derived from a survey of 11 stomal therapy nurses internationally (including one Australian) and it was assumed that the same management algorithms would be applicable to Australia; and
3. Costs were likely to be inflated particularly with regard to the choice of MBS item fora stoma nurse visit. A visit was costed at $55.85 which is the Medicare scheduled fee for a nurse practitioner visit of greater than 40 minutes duration. This seemed excessive, especially for the management of mild or moderate skin disease. Other assumptions made indicated the likely inflation of modelled treatment costs, including that 100% of cases of moderate/severe PSCs would require treatment with infliximab, and that a proportion of patients would require four nurse and four specialist visits over a seven week period.

Reduction in Accessory Use

The Panel considered the applicant’s calculation of the total premium component due to the reduction in the use of accessories observed in the Dialogue Study (also divided by the number of trial units based on SAS maximum quantity thresholds) to be unsound, given that this would also have been highly influenced by nurse interaction at visit one and was therefore also likely to be highly confounded. It was the opinion of the Panel that if people were using accessories to provide greater protection/security/comfort etc. at study entry, they would be unlikely to change this in the short-term unless it was advised or suggested by nurses that they do so due to a belief that the SenSura product was superior. The Panel noted that this issue was not addressed in the submission.

Willingness to Pay

The Panel noted that the applicant presented two documents supporting willingness to pay – the first an unpublished French study and the other a publisheddiscrete-choice experiment by Bonnichsen (2011). The French study reported a willingness to pay among users for the flexibility and comfort, performance with regard to skin condition and leakage, and filter performance of stoma appliances. Bonnichsen reported that respondents had a significant positive willingness to pay for all potential attribute improvements, with the strongest preference being exhibited for a reduction in the incidence of leakage and improvements in system flexibility.

The attributes for which the applicant sought an incremental value based on willingness to pay in this submission were (the superior performance of the SenSura product in relation to) skin condition, the related problem of leakage and filter performance. As previously stated, the Panel considered the evidence of improved skin condition outcomes due to the use of the SenSura appliance to be highly confounded. The Panel then noted that the frequency of leakage reported in the Dialogue Study was likely to be confounded by the nursing effect, even though it was possible to determine extent. Filter lifetime was assumed by the applicant to be longer and to represent a value. Due again to issues ofthe quality of clinical evidence, the Panel considered any value increment to be highly uncertain.

The willingness to pay component of the per unit price premium was calculated per month (as opposed to per trial period), by dividing the total value claimed by the maximum monthly quantity of products subsidised under the SAS. The Panel noted that this method of distributingthe aggregate price premium requested across a quantity unitsbased on SAS maximum quantity thresholds, as was applied in all three components of the applicant’s economic case,resulted in the applicant’s requesting widely varying unit price premiums for the same feature across multiple submissions – the greatest being for products subject to the lowest maximum monthly quantities. In a context in which the SPAP has previously recommended absoluteunit price premiumsfor features demonstrated to improve health outcomes,it was noted that this methodology was inappropriate.

Financial Analysis

Not undertaken.

Overarching Summary of SPAP Position