Article Text

Going ‘Super Value’ in New Zealand: cigarette pricing strategies during a period of sustained annual excise tax increases
  1. Philip Gendall1,
  2. Katherine Gendall2,
  3. J Robert Branston3,
  4. Richard Edwards1,
  5. Nick Wilson1,
  6. Janet Hoek1
  1. 1 Public Health, University of Otago Wellington, Wellington, New Zealand
  2. 2 Independent Researcher, Wellington, New Zealand
  3. 3 School of Management, University of Bath, Bath, UK
  1. Correspondence to Professor Janet Hoek, Public Health, University of Otago, Wellington, P O Box 7343, New Zealand; janet.hoek{at}


Background Between 2010 and 2020, the New Zealand (NZ) Government increased tobacco excise tax by inflation plus 10% each year. We reviewed market structure changes and examined whether NZ tobacco companies shifted excise tax increases to maintain the affordability of lower priced cigarette brands.

Methods We cluster-analysed market data that tobacco companies supply to the NZ Ministry of Health, created four price partitions and examined the size and share of these over time. For each partition, we analysed cigarette brand numbers and market share, calculated the volume-weighted real stick price for each year and compared this price across different price partitions. We calculated the net real retail price (price before tax) for each price partition and compared these prices before and after plain packaging took effect.

Results The number and market share of Super Value and Budget brands increased, while those of Everyday and Premium brands decreased. Differences between the price of Premium and Super Value brands increased, as did the net retail price difference for these partitions. Following plain packaging’s implementation, Super Value brand numbers more than doubled; contrary to industry predictions, the price difference between these and higher priced brands did not narrow.

Conclusions Between 2010 and 2020, NZ tobacco companies introduced more Super Value cigarette brands and shifted excise tax increases to reduce the impact these had on low-priced brands. Setting a minimum retail price for cigarettes could curtail tobacco companies’ ability to undermine tobacco taxation policies designed to reduce smoking.

  • price
  • public policy
  • taxation
  • packaging and labelling

Data availability statement

Data are available in a public, open access repository. Data are available from the NZ Ministry of Health Tobacco Returns website.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


  • Tobacco companies can undermine tobacco excise tax increases designed to prompt quitting and discourage smoking uptake by disproportionately increasing the price of Premium brands (over-shifting) while minimising the price increases applied to Budget brands (under-shifting).


  • Between 2010 and 2020, New Zealand-based tobacco companies used differential price shifting to reduce the impact of annual tobacco excise tax increases on lower priced brands compared with higher priced brands.

  • Following the implementation of plain packaging, tobacco companies introduced several new Super Value brands; while the market share of these brands increased, the price difference between lower priced and higher priced brands did not narrow.


  • Setting a minimum retail price could prevent tobacco companies from undermining the intent and impact of tobacco excise tax increases.


Marketers manipulate prices to foster brand switching, prompt trial and impulse purchase, and maintain purchase patterns1; governments have responded by increasing tobacco excise taxes to decrease youth uptake and adult smoking prevalence.2 In high-income countries, for every 10% tobacco excise tax increase fully applied to tobacco prices, consumption typically falls by 4%–5% as rising prices stimulate quit attempts and reduce smoking uptake.3–5

Industry documents reveal that tobacco companies used pricing strategies to shape the behaviour of people who smoke and foster uptake among young people.6 They resist excise taxes by arguing these are regressive, compound disadvantage,7 and foster trade in illicit tobacco.8 Nonetheless, independent analyses have found that excise tax increases may reduce health inequities; they also reveal tobacco companies’ involvement in illicit trade and present strategies to reduce illicit tobacco supply.3 9–11

Tobacco companies undermine excise tax increases by disproportionately passing these onto Premium brands (over-shifting) and away from Budget brands (under-shifting).12 This strategy reduces the incentive for people on lower incomes to quit smoking in response to rising prices. A systematic review examining excise tax increases in Australia, Bangladesh, Taiwan, Spain, the USA and European Union countries found strong evidence of tax over-shifting and under-shifting.13 A Canadian study not included in the review found that tobacco companies responded to excise tax increases by supplying lower priced cigarettes to people who might otherwise have quit, but maintained their revenues by over-shifting the taxes to more expensive brands.14

An earlier UK study of cigarette prices between 2000 and 2009 found that the real weighted average price of Premium, mid-priced and economy brands increased gradually over the period concerned, while the real price of ultra-low-priced (ULP) brands barely changed.15 Between 2006, when ULP brands first appeared in the UK, and 2009, the average price of ULP brands increased by just 1.3 pence per stick compared with increases of 4.1–4.9 pence for the more expensive brands. Because taxes were under-shifted in the ULP partition and over-shifted in the other three partitions, the real price gap between these market partitions widened.15

A subsequent UK replication study using data from 2009 to 2015 produced similar results; despite regular tax increases, average real prices for the cheapest cigarette partitions remained steady from 2013, while sales volumes grew.16 Reductions in pack size and price-marking (printing the recommended retail price (RRP) on packs) maintained low prices. Although the authors noted that plain packaging would prevent both strategies, they called for further changes in tax policy to minimise tobacco companies’ attempts to thwart government health policy.16

Sheikh et al 13 found that, as well as differentially shifting taxes between brands, tobacco companies reduced the impact of tobacco excise increases by launching new, low-priced brands using price smoothing (ie, reduced the impact of price increases by introducing these gradually rather than as one sharp increase) and price discrimination (ie, targeted pricing to specific consumer groups), developing product promotions and changing product attributes.13 However, in high-income countries, tobacco companies often over-shifted excise tax increases to higher-priced products.15

We examine tobacco companies’ use of these strategies in New Zealand (NZ), where from 2010 to 2020, the government applied annual excise tax increases of at least 10% to all tobacco products. The tax was a fixed amount per stick and the annual adjustment also included an inflation uplift measured by the Consumers Price Index (CPI) in the previous year; the tax was applied on 1 January each year (with the final increase applied on 1 January 2020). We also examined pricing responses to plain packaging, introduced in March 2018 (with full implementation 3 months later).

To date, few studies document how tobacco companies respond to sustained excise tax increases. A 2014 study compared the RRP of three British American Tobacco (BAT) cigarette brands (representing the Premium, mainstream and Budget price partitions) with these brands’ expected retail price, calculated by applying the annual excise tax and CPI increases introduced by the NZ Government in 2010.17 This study found the median increase in price before and after the 2014 tax increase of 10% was only 3% for the Budget brand, but 8% for the Premium brand and 11% for the mainstream brand. The authors concluded that tobacco companies were undermining the excise tax policy and impeding realisation of the NZ Government’s Smokefree 2025 goal.17 However, the three brands analysed may not have represented the total market. Further, the study examined prices in relation to a single tax increase and did not assess how tobacco companies use pricing to respond to policies such as plain packaging, which reduces the value proposition tobacco brand imagery represents.18 19

Tobacco companies have argued that plain packaging would reduce brand differentiation and consumer choice, and lower prices across market partitions, thus likely increasing smoking prevalence.20 UK studies undertaken after plain packaging was implemented in May 2017 found no evidence prices had reduced over the long term or that the price range had narrowed.16 21 Instead, they reported reduced under-shifting, although this change occurred alongside introduction of a minimum excise tax.16 21 A post-implementation study in Australia, the first country to introduce plain packaging, found that prices displayed on price boards did not decrease post-plain packaging.22 Analyses of RRPs post-plain packaging also found these were higher in real terms (ie, increased beyond inflation) and continued to rise after plain packaging.23 24

Australian studies noted the growth of a Super Value market partition post-plain packaging.25 A large survey of Australians who smoke reported use of value brands had increased and found that price increases post-plain packaging occurred in all market partitions, particularly the Premium partition.26 More recent work found tobacco companies protected the new value partition they had created by cushioning tobacco tax increases and passing these on incrementally over several months.24

Because NZ now has among the highest tobacco prices of any country, tobacco companies have a strong incentive to undermine tobacco excise taxes (the 2021 price of a pack of 20 Marlboro cigarettes was NZ$36.90 (~US$25)). NZ requires tobacco companies to furnish annual data on the number of cigarettes released for each brand, variant and pack size, and RRPs. These data offer a unique opportunity to gain insights into how a tobacco market evolved over an extended period of excise tax increases and following implementation of plain packaging. Given the earlier NZ study17 and international findings,13 15 16 we hypothesised that the number of Budget brands and their market shares would increase, and that the price differential between Budget and Premium brands would also increase.


We sourced the annual manufactured cigarette returns between 2010 and 2020 for BAT, Imperial Tobacco (IMP) and Philip Morris International (PMI) from the NZ Ministry of Health website (these companies account for 99% of the tobacco sold in NZ).27 The annual returns report the price and volume of sticks released for sale for every cigarette brand imported into or manufactured in NZ. Price discounting promotions are not permitted in NZ, but some small retailers may have charged more or less than the RRP for cigarettes they sold. However, Marsh et al estimated that 82% of the small retailers they surveyed in 2013–2014 complied with BAT’s recommended RRPs17; consequently, the prices reported in the annual cigarette returns represent the retail prices paid by most NZ people who smoke. Between 2010 and 2020, manufactured cigarette sales accounted for between 64% and 73% of tobacco sales volume in NZ (sales of roll-your-own (RYO) tobacco accounted for nearly all the remainder). Because the RYO market has fewer brands, saw market share in 2020 increase by only four percentage points compared with 2010 and had a largely constant price relative to Super Value cigarettes between 2017 and 2020, we focus on manufactured cigarettes.

Creating price partitions

We created price partitions by performing cluster analysis that allocated brands to a specific partition for each year of our study. We applied the RRP for 20-stick packs, which constituted 53% of all packs sold between 2010 and 2020 (range: 44.2% in 2017 to 65% in 2020) to all pack sizes of that brand (the other main pack sizes were 25 and 30). This process standardised the price for each pack size and brand, removed the volume discount effect of bundled packs and packs with a ‘free’ cigarette (eg, 21 cigarettes in a notional 20-stick pack) and allowed us to include all the sales data in the price partition analysis. (Volume discounts would have reduced the unit stick price for the brand but conceptually would not have changed the price partition of the brand; see online supplemental file 1 for further information.)

Supplemental material

We normalised the standardised cigarette price data for each year using min-max scaling to remove the yearly price increase effect (ie, converted all prices in each year to values between 0 and 1 to remove the effect of prices increasing over time).

Embedded Image

where: z it =scaled price for brand i in year t; x it =original price for brand i; min (x t )=the lowest price for any brand in year t; and max (x t )=the highest price for any brand in year t.

We cluster-analysed the entire normalised data set using the R package Ckmeans.1d.dp and performed univariate k-means clustering to estimate three price clusters for the total data set: Budget/value, Everyday and Premium.28 29 We then separately analysed the Budget/value cluster into two clusters: Budget and Super Value (the data for this latter analysis were re-normalised against the Budget/value data). We also estimated a four-cluster solution directly, but the goodness of fit was lower than for the two-stage clustering method described. We assessed the face validity of the clustering results and reallocated a small number of brands to correct marginal partition anomalies (see online supplemental file 1 for further details).

Price analyses

Using the RRPs for each brand and pack size (ie, unstandardised prices), we conducted two price analyses. First, we divided the RRP for each brand by the pack size to calculate the unit stick price. We then divided the unit stick price by the CPI (using 2010 as the base) to remove the impact of inflation and hence produce the real unit stick price for each year (see online supplemental file 2). We multiplied the real unit stick price by the sales volume for each brand and then aggregated the results across brands to calculate the mean volume-weighted real stick price for each price partition in each year. From these prices, we calculated the absolute price difference between the mean volume-weighted real stick price for the Premium partition and the Super Value partition.

Supplemental material

For the second price analysis, we calculated the inflation-adjusted (real price) mean net retail price for each price partition in each year by subtracting the retail sales tax (in NZ applied to all goods and services at 15%) and excise tax from the mean real RRP for all brands in a partition and adjusting the result for inflation. In 2010, two excise tax levels applied; we used the 29 April excise tax (ie, the tax that applied for most of the year). We examined changes in the net retail price to identify pricing behaviours and interpreted an increase in net retail price as over-shifting and a reduction as under-shifting.

Sales volume market shares and brand numbers

We also calculated the volume market share for each price partition in each year and recorded the number of brands and pack sizes within each partition for each year.


Between 2010 and 2020, BAT, IMP and PMI sold approximately 50 different brands in NZ (‘brands’ refers to overall brands, such as Pall Mall, rather than brand variants, such as Pall Mall Red, Pall Mall Blue and Pall Mall Green); around a quarter of these were sold in each of the years analysed. Brand numbers in the Super Value and Everyday partitions were relatively stable until 2020; Super Value brand numbers ranged from three to five, before increasing to seven in 2020. Budget brand numbers ranged from five to seven (six in 2020) and Everyday brand numbers ranged from nine to 12, before declining to six in 2020. Premium brand numbers remained at 12–13 until 2014 when they declined to eight; from 2016 to 2019 the Premium partition contained five brands, though numbers increased to eight in 2020. By 2020, Super Value brand numbers had more than doubled while Premium brand numbers had nearly halved (see online supplemental table 2). The period examined thus saw the composition of price partitions change markedly as Super Value brands proliferated and Premium brands contracted.

Table 1 shows the volume share of each market price partition over time. Between 2010 and 2020, the market shares of Super Value and Budget brands increased markedly, while those of Everyday and Premium brands decreased. The increase in Super Value and Budget brands’ market share was particularly marked after plain packaging was fully implemented in June 2018. Specifically, Super Value brands’ market share increased from 3.2% in 2010 to 24.4% in 2020 and Budget brands’ market share increased from 16.8% in 2010 to 49.6% in 2020. Everyday brands’ market share decreased from 52.6% in 2010 to 15.2% in 2020, while Premium brands’ market share decreased from 27.3% in 2010 to 10.7% in 2020.

Table 1

Price partition market share of manufactured cigarette brands in NZ: 2010–2020

Table 2 shows changes in volume-weighted real stick prices over time. Between 2010 and 2020, the absolute difference between the price of Premium brands and Super Value brands increased from 19.7 cents per stick to 30.5 cents per stick. The differences were greater in 2018–2019, immediately following plain packaging, but reduced slightly in 2020. Between 2010 and 2019–2020, real unit stick prices increased more for Everyday (87.4 cents) and Premium (86.7 cents) brands than for Budget (77.7 cents) or Super Value (75.9 cents) brands.

Table 2

Volume-weighted real stick price by manufactured cigarette brand partition: 2010–2020

Figure 1 illustrates trends in volume-weighted real stick price for each price partition in table 2 and shows that prices did not reduce following plain packaging’s introduction. Real prices in all price partitions increased markedly in 2018; while these flattened off in 2020, they were at least 25% higher in 2020 than in 2017.

Figure 1

Average volume weighted real price by price partition: 2010–2020

We next analysed net real retail prices. If the excise tax had been fully passed on to the observed retail price, the net inflation-adjusted price in each price partition would remain constant between years. However, a decrease in the net price from year to year suggests the excise taxes were under-shifted, or, conversely, over-shifted if the net price increased. Table 3 highlights the net real prices that declined or increased year on year.

Table 3

Net real cigarette prices for manufactured cigarettes by market partition of brand: 2010–2020

Our findings indicate consistent over-shifting of excise tax on brands in all price partitions for most years between 2011 and 2020 where the year-on-year change was one cent or more (we consider a change of less than one cent as effectively no change). Comparing the mean differences in net prices for each year to the net price in 2010 shows that the mean difference for Super Value brands and Budget brands was an increase of 8.0 and 9.5 cents, respectively; those for Everyday and Premium brands were 13.7 and 15.7 cents, respectively. Over-shifting for more expensive brands was nearly double that for lower priced brands, increasing the differential in price between the price partitions. Figure 2 illustrates the patterns of these changes in net prices and shows that net real prices for Everyday and Premium brands increased more sharply than prices for Super Value or Budget brands over the period concerned.

Figure 2

Net real stick price by price partition: 2010–2020

In 2019, following full implementation of plain packaging, the real net price of Super Value brands decreased by 8% and fell further in 2020. Similarly, the real net price of Budget, Everyday and Premium brands declined by around 5% between 2018 and 2020, and plain packaging was associated with excise tax under-shifting in all price partitions.

As well as price shifting across brands over time, tobacco companies could potentially manipulate pack sizes within brands to under-shift excise tax increases; we undertook a preliminary analysis of this question. Between 2010 and 2018, pack sizes ranged from 20 to 60 sticks; pack size proliferation was particularly marked in the Budget and Everyday partitions (though 20 packs were almost always the single biggest pack size category) (see online supplemental file 3). The unit stick prices for larger packs were sometimes lower than the unit stick price for a pack of 20, though not always. Where larger packs offered a ‘quantity discount’, doing so effectively meant under-shifting the excise tax on these packs because this tax is levied per stick. For example, in 2017, the price of a Horizon 20-packwas $24.50 and the price of a Horizon 21-pack was $25.00, a difference of 50 cents for one extra stick when the excise tax was 73.8 cents per stick and the per-stick price was 122.5 cents.

Supplemental material


To discourage smoking uptake among youth and encourage cessation, the NZ Government raised tobacco excise taxes between 2010 and 2020, making the real price of cigarettes among the highest globally. Our analyses suggest tobacco companies undermined this policy by reducing Everyday and Premium brand numbers while proliferating Budget and Super Value brands. For example, in 2019 and 2020, tobacco companies introduced three low-priced variants of existing brands—Rothmans Royals, Rothmans London and Marlboro Crafted—and reintroduced Chesterfield. These additions more than doubled the number of Super Value brands available, giving people who smoke more lower priced brand options.

Adding new brands, particularly those with sophisticated names that did not have pre-existing low-cost connotations, potentially ameliorated negative perceptions of trading down and facilitated movement to cheaper brands. Individually, these new brands had relatively small market shares, but growth in this price partition meant the overall market share of Super Value brands increased from 3.2% to 24.4% over the study period. Even more striking was the increase in market share of Budget brands (from 16.8% in 2010 to 49.6% in 2020); overall, the market share of brands in these two partitions increased from 20.0% to 74.0%. Despite tobacco companies’ predictions that plain packaging would narrow the price range between lower priced and higher priced brands,20 30 31 the price difference between Super Value and Premium brands widened in both 2019 and 2020.

Between 2010 and 2018, tobacco companies appear to have over-shifted the annual excise tax increases in all price partitions in most years, though over-shifting was greater for Everyday and Premium brands than for Super Value or Budget brands. Differential shifting of tax between lower priced and higher priced brands meant Super Value and Budget brands became relatively cheaper compared with Everyday and Premium brands; this differential increased over time. Furthermore, though many year-on-year changes in net real prices were small, because these were spread over hundreds of thousands of cigarette sticks, they had considerable revenue, and potentially profit, implications.

Our results suggest plain packaging may have subsequently led to undervshifting of the excise tax increases in NZ, particularly for Super Value and Budget brands. In 2019, the mean real net price of Super Value brands decreased, and in 2020 the real net price for all price partitions decreased (though actual RRPs continued to increase in all price partitions). Decreases in net real prices may indicate that excise tax had increased to a point where the industry had to change its pricing strategy to one of consistent under-shifting. Alternatively, these decreases could reflect plain packaging’s effects, which reduced opportunities to use branding to support differential prices. Our findings are consistent with price-shifting practices observed in the UK and elsewhere,13 16 and provide additional evidence to challenge pricing arguments used to deter the introduction of plain packaging.21

Introducing a minimum price alongside plain packaging could reduce tobacco companies’ ability to under-shift tobacco excise tax increases. For example, an American study modelling different minimum retail price scenarios concluded that minimum pricing could be very effective if it was sufficiently high; importantly, it could also promote health equity.32 Simulation studies undertaken in California predicted minimum retail prices would lead to substantial decreases in smoking prevalence and smoking intensity,33 34 with the greatest reduction in smoking prevalence occurring among young people and under-privileged populations.34 A survey of responses to a hypothetical minimum retail price for tobacco among people who smoke in the UK provided further support for a minimum price strategy. Approximately 20% of those surveyed indicated they would smoke less or quit, and around 40% indicated that minimum retail prices would help them to stay quit.35

Other product categories, such as alcohol, indicate the political feasibility and effectiveness of minimum pricing policies. In May 2018, Scotland introduced a minimum unit price for alcohol to reduce the availability of very cheap, strong alcohol; Wales followed suit in March 2020. Evaluations show subsequent increases in the average price of alcohol and reduced household purchases of alcohol in both Scotland and Wales, compared with England (which had no minimum retail price).36 Ribisl and colleagues recently argued that a minimum tobacco price law, together with a floor price, could reduce under-shifting and that regular increases in the minimum tobacco price could eventually eliminate Super Value and Budget price partitions.12

While a minimum retail price would likely reduce tobacco consumption, revenues from this strategy would accrue to retailers and manufacturers, rather than to the government. Although tobacco companies are thus less likely to oppose this measure (relative to excise taxes), few policy makers would wish to enhance the tobacco industry’s profits. Consequently, some tobacco control advocates have proposed setting industry pre-tax prices, to which they propose adding excise taxes, or establishing a minimum excise tax.12 37 38 For example, Scollo and Branston propose setting a maximum wholesale price alongside excise tax increases, where any increased industry profits would be paid to the government.39 Moving control over pricing from the industry to government would also enhance monitoring of industry cost structures and build on licensing requirements, such as imposing conditions on tobacco sales (eg, limiting pack numbers sold in a single transaction).39 Regulation of tobacco prices could thus prevent undershifting, while also reducing tobacco companies’ ability to profit from the policy.

Setting minimum retail prices or excise tax would give effect to government policy, but would affect people on lower incomes, among whom smoking prevalence is greater.40 Policies providing enhanced, targeted support to people who smoke should therefore accompany measures that increase tobacco costs. Such an approach would increase self-efficacy and successful cessation among the many people who regret ever having started smoking,41 and respond to calls for support that people from these communities have made.42 43

Like all studies, ours has strengths and weaknesses. As explained in online supplemental file 1, inconsistencies in the data analysed required us to make judgements; we have documented and justified our approaches but had no control over the data quality. Our approach to identifying price partitions differs from the methodology employed in studies using tobacco companies’ descriptions of their brand profiles. Our analyses were based on one reported price per year; because tobacco companies may change prices several times in a year, we may not have detected excise tax price shifting within years. However, our price partitions have strong face validity and the patterns identified are consistent with those found in other high-income countries. Furthermore, because NZ’s tobacco control legislation prohibits price discount promotions on tobacco products, we believe the prices analysed accurately reflect retail prices and provide robust insights into tobacco price trends in NZ. However, our findings give rise to new questions, such as whether and how price manipulation occurs within as well as between brands; future research could explore this question.

Both centre-right and centre-left administrations have used regular, large increases in tobacco excise taxes to progress the Smokefree 2025 goal. However, tobacco companies have undermined this strategy by introducing new Super Value and Budget brands, by pricing strategies that increased the differentials between these price partitions and Everyday and Premium brands, and by under-shifting of excise tax increases post-plain packaging. These measures increased the market share of cheaper brands and dampened the impact of tobacco excise and plain packaging policies, thus undermining the goal of reducing smoking prevalence. Although a discussion document setting out proposed measures to achieve NZ’s Smokefree 2025 goal recognised minimum retail pricing could reduce price manipulation, the final Action Plan did not include a minimum retail pricing strategy.44 45 Our findings suggest including a minimum pricing policy in NZ’s proposed legislation would close a long-standing policy loophole and counter tobacco companies’ attempts to thwart government policies designed to protect public health.

Data availability statement

Data are available in a public, open access repository. Data are available from the NZ Ministry of Health Tobacco Returns website.

Ethics statements

Patient consent for publication


Supplementary materials


  • Twitter @JRBranston

  • Contributors PG and JH conceptualised the study. PG and KG undertook the data analyses. PG and JH wrote the manuscript. JRB advised on analyses and, with RE and NW, commented on advanced drafts. PG and JH led the revisions with input from all authors. All authors have approved the final manuscript. PG is the guarantor.

  • Funding PG, JH, RE and NW are funded by the University of Otago, NZ. KG is an independent researcher. JRB is supported by Bloomberg Philanthropies Stopping Tobacco Organizations and Products project funding (

  • Competing interests JRB owns 10 shares in Imperial Brands for research purposes. The shares were a gift from a public health campaigner and are not held for financial gain or benefit. All dividends received are donated to health-related charities, and proceeds from any future share sale or takeover will be similarly donated.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.