Non-PHARMAC medicines coverage – aiming for a data-informed evaluation

Non-PHARMAC coverage has become increasingly important to clients purchasing private health insurance, primarily due to rising public demand for treatments and medications not funded by PHARMAC (the government agency overseeing public pharmaceutical purchasing and subsidy).

Wednesday, April 23rd 2025, 11:57AM

by Russell Hutchinson

Non-PHARMAC coverage has become increasingly important to clients purchasing private health insurance, primarily due to rising public demand for treatments and medications not funded by PHARMAC (the government agency overseeing public pharmaceutical purchasing and subsidy).

Recent coverage focused on the severity of a serious illness and whether it can be cured as a potential limitation to non-Pharmac coverage. We felt it was worth clarifying that, in effect every policy with a non-PHARMAC benefit includes some limitations of this nature – though they may appear in different places and be expressed in different ways within the policy wording. Every private medical insurance contract limits access to treatment which could, in effect, last a long time and not get anywhere – geriatric care, chronic conditions, treatments that are not medically effective, exclusions, or requirements that the condition must be medically significant, no coverage for non-surgical hospital care, these are all forms of limitations (present in all the policies we review) that limit coverage around this issue.

They appear in different places – such as the exclusions, the benefit definition, and the terms. We consider the requirement for ‘significant impact’ to be equivalent to ‘serious or imminent threat to life.’ That’s based on both claim evidence and legal case findings in court cases where there was an attempt by a client to claim that ‘terminally ill’ was equivalent to a threat to life and the judge ruled in favour of the insurer. Other limitations appear by omission – for example, there being no coverage for non-surgical hospital admissions in some contracts. This is because medical insurance contracts are not intended for or priced to cover, in effect, old age. That coverage is known as long-term care insurance, and it is not currently available in New Zealand.

While we can construct possible claim scenarios and thought experiments to help us explore issues and assess them against policy wording, these are less useful as analytical tools than we would like – especially if they are non-specific and not based on actual conditions, but purely hypothetical. When considering a scenario, we prefer to look up actual New Zealand incidence rates for disorders rather than focusing on unnamed and unquantified ones. What’s most important is a value-based measure - one that considers real conditions or disorders that are covered or not, the most common claims in the category, whether cover extends to those areas, the typical cost of treatment, and which claims are commonly accepted or declined. Although insurers are sometimes less forthcoming than we would like, partly because of historical challenges with claims data recording, much can be learned with a dedication to examining other sources of data: the ministry of health publishes extensive disease incidence and hospital admission statistics, for example.

Providers of private care are often happy to share numbers too, including costs of treatment. With these data we can ask better questions and weight impacts appropriately. Data is our preferred approach when assessing ‘I reckon’ statements. We like to reframe these as questions – they can be a valuable starting point, but rarely fully quantify the issue sufficient for rating purposes.

What to consider in evaluating non-PHARMAC coverage

At Quality Product Research (Quotemonster), we aim to simplify the complexity surrounding non-PHARMAC coverage. Last year, we created a reference guide for our subscribers to quickly identify key differences and limitations across medical insurance products. Not all non-PHARMAC coverage is equal, and while it's beneficial that some insurers include this coverage, limits are often restrictive. The most important limits are those which restrict the response of the product in dollar terms – the sub limit applied to the benefit – and whether it is limited to cancer care or includes other conditions. Beyond that we get into the area of whether the medicine must be used in accordance with Medsafe indications.

Not all medicines have been approved for all possible uses in New Zealand – that doesn’t mean that it isn’t used for that purpose in other countries. It doesn’t mean that it’s unsafe, only that approval has not yet been sought and received in this market. Often that can be as much for economic reasons as clinical ones: we are a small market, and getting a new medicine approved in, for example, the United States or Germany, will be priorities for a pharmaceutical company that is eager to recover the very substantial development costs of a new medicine. Recent news (see link below) suggests that approvals should be quicker in future, due to the ‘rule of two’ being proposed, perhaps reducing the impact of this restriction, but until then we recognise this difference.

Once again, difficult decisions must be made because any model is necessarily a simplification of reality. Rating based on all potential real-life outcomes is impractical for any rating agency. While our model cannot reflect all the complexities inherent in medical insurance, our objective is to use logic and reason to highlight the relative value of benefits paid. By utilising claims data provided by reinsurers and insurers, we construct realistic claim scenarios reflecting typical coverage scopes.

In our medical claims model, we assume a policyholder will purchase health insurance at age 39 (a median supported by insurer and quote data), maintain this policy for 18 years, and incur approximately $100,000 in claims over this period (though actual claims vary significantly). Unsurprisingly, non-PHARMAC coverage introduces additional complexity, and our model cannot encompass every possible circumstance. We model a 10% likelihood over 18 years that a non-PHARMAC claim will occur (incidence), and when such a claim occurs, it will be fully paid once (100% frequency). Our scoring aims to reflect typical claim amounts, acknowledging the significant variations between insurers while accurately capturing the relative differences to highlight the differences in available response under each product.

Non-PHARMAC coverage is categorized into two distinct items: Non-PHARMAC medicines (cancer-specific) and Non-PHARMAC medicines (non-cancer), and within these we consider sub items that deal with factors such as: limitations on in-home coverage, whether the insurer provides coverage beyond chemotherapy or immunotherapy treatments, whether hospitalisation is a claim criterion, and whether there is coverage for Medsafe-approved medications used outside their approved indications.

Access to new drugs to speed up if approval given by some countries | RNZ News

Tags: Medsafe Pharmac Russell Hutchinson