A Survival Gap That Was Hidden by a Checkbox

Interview and article by Pamela Pont

For years, Native Hawaiian and Pacific Islander women were counted together with Asian women in US cancer statistics. When Dr. Melissa Merritt's team separated them across 18 studies, a survival disparity in ovarian cancer appeared.

The paper: Meagher NS, White KK, Wilkens LR, et al., Merritt MA. "Differences in overall survival for invasive epithelial ovarian cancer by race and ethnicity: results from the Ovarian Cancer Association Consortium." British Journal of Cancer, June 2026. DOI: 10.1038/s41416-026-03509-8

The researcher: Dr. Melissa Merritt, senior author, The Daffodil Centre, The University of Sydney and Cancer Council NSW.

The short version

Ovarian cancer survival is not the same for everyone, and researchers have known for some time that Black women in the United States fare worse than Non-Hispanic White women. What has been much harder to see is what happens to smaller groups, because there has not been enough data on them to study.

Dr. Merritt's team pooled data from 18 studies in the Ovarian Cancer Association Consortium, a shared resource built over roughly 15 years by investigators who agreed to combine their individual studies into one pool. That gave them over 14,600 women with invasive epithelial ovarian cancer, and enough numbers to look at groups that usually disappear into a larger category.

Compared with Non-Hispanic White women, Native Hawaiian and Pacific Islander women had about 1.6 times higher risk of death, and Black women about a 1.3 fold higher risk. Among women with high-grade serous carcinoma, the most common and aggressive form of the disease, the gap for Native Hawaiian and Pacific Islander women was wider still: roughly twice the risk.

These differences held after accounting for age, year of diagnosis, disease type and stage, body mass index, smoking, and postmenopausal hormone use. Which is to say the gap is not explained by those factors, and the study cannot yet say what does explain it.

Why separating these groups mattered

This finding exists because of a decision about categories.

Native Hawaiian and Pacific Islander people have historically been counted together with Asian people in United States census and health data. More recently, that has begun to change, and researchers have started reporting on the groups separately.

That distinction is what made this result visible. Grouped together, the numbers averaged out. Separated, a disparity emerged that was not appreciated in ovarian cancer before.

"When Asian and Pacific Islander women were grouped together, it was hiding the worse survival outcomes for the Pacific Islander women," Dr. Merritt says.

The team had reason to expect it. Poorer health outcomes for Native Hawaiian and Pacific Islander communities are well documented across other cancers and chronic diseases, along with a younger age at diagnosis. What was missing was evidence in ovarian cancer specifically, and the sample size to produce it.

How the study was built

Dr. Merritt had asked a version of this question before, in a United States military health system population. That study drew on cancer registry data, which meant it captured who was diagnosed and what happened, but almost nothing about the women themselves.

"We didn't actually know anything about any of the lifestyle factors or the things that we measure on the questionnaire," she says.

It also had a feature that made it unusual: in that system, everyone has or should have equal access to health care. That removes access as a variable, which is scientifically useful but also means the study could not speak to what happens in populations where access differs.

The consortium data offered the opposite trade. Questionnaire responses meant the team could account for smoking, body mass index, and postmenopausal hormone use, factors known to influence ovarian cancer survival. And because access to care varies across these populations, the results may reflect it, which is part of what makes them worth pursuing.

The consortium itself is the quiet enabler here. Individual investigators, each with their own case-control study, pooled their data into a shared resource that other researchers can apply to use. Dr. Merritt is careful about her role in it.

"I've been the beneficiary of having access to the data and developing a project and being able to conduct the study," she says. "I wasn't directly involved in setting everything up. So I think that just goes to show that it's been a really valuable resource that people have set up, and it's available for researchers."

What this means, and what it does not

A pattern, not an explanation. This is the point Dr. Merritt returns to most often. The study establishes that a survival gap exists. It does not identify the cause.

"We identified a pattern, but we don't know the reasons for it," she says. "The study value is highlighting that this is an issue, but really it's highlighting the issue so that further work can be done to correct these disparities."

What the study could not account for. The team had no information on the treatment each woman actually received, or whether it met clinical guidelines. That is a substantial gap when the question is why outcomes differ, and Dr. Merritt names it as the limitation she would most want to fix.

Almost all of the Native Hawaiian and Pacific Islander data came from Hawaii. With 98 women in that group, drawn largely from the Hawaii Ovarian Cancer Case-Control Study, the findings do not tell us about Pacific Islander communities in the mainland United States. Dr. Merritt notes there is currently no population-based study that would allow that question to be asked in ovarian cancer at all.

No differences in modifiable factors were found. The team tested whether things like smoking and body mass index related to survival differently across racial and ethnic groups. They did not find such differences, and the sample sizes outside the White participant group were too small to draw firm conclusions either way.

Dr. Merritt had hoped for something actionable here. "It would be nice to make a recommendation like that, but that actually didn't come out of this study."

What this does not mean for an individual. These are group-level averages across thousands of women, not a prediction about any one person's outcome. A hazard ratio describes a pattern across a population; it does not tell a woman what will happen to her. Nothing in this paper changes what any individual patient should do, and the study identifies no different guidance for anyone based on their race or ethnicity. General survivorship advice that applies broadly, including not smoking, still applies here.

What's next

Finding a population where the question can be asked. The most concrete barrier is that the data does not exist. Dr. Merritt is presenting this work at the Australasian Epidemiological Association meeting in Fiji in October, partly to find out whether other Pacific Islander populations could be studied. "I don't know of anyone really working in ovarian cancer in that space," she says.

Treatment data. Future studies of this kind need information on what care patients actually received, so the question can move from whether a gap exists to why.

Broader diversity in study populations. Her answer to what she would do with unlimited funding was not a new analysis but new data: studies designed from the start for racial and ethnic diversity, so that subgroups currently too small to study become large enough to learn from.

Biology, alongside epidemiology. She points to work examining patients' tumor tissue to understand whether the disease itself differs across groups, which she sees as a necessary complement to the population-level picture.

Open to collaborate

Dr. Merritt is actively looking for ways to study ovarian cancer outcomes in Pacific Islander populations beyond Hawaii, and is not aware of others working in that space. Researchers with access to relevant cohorts, registries, or population-based studies, particularly in the Pacific or in mainland US Pacific Islander communities, would be worth connecting with her.

In her own words

What drew you to ovarian cancer epidemiology?

I was interested in studying the whole person and why people get cancer. That drew me to epidemiology; to do work on a population basis and look at the factors we're exposed to every day and how that might relate to risk of developing cancer.

I've always focused on ovarian cancer. Since I started working on cancer, it's been ovarian cancer, because my grandmother had ovarian cancer and I realized I didn't know very much about it. So, it's because of that personal connection, and wanting to work on women's health as well.

How would you explain the main finding to a patient rather than a scientist?

For each comparison we have a reference group, and because the White women in the study are the largest group, we compare with those women. So you would say that if you are a Native Hawaiian or Pacific Islander woman diagnosed with ovarian cancer, your risk of mortality is about 1.6 times higher compared to a White woman. We also saw a disparity for Black women, where they had a 1.3 times higher mortality, confirming previous results.

What would you want a clinician to do differently after reading this?

Hopefully they would at least recognize that race and ethnicity does impact survival outcomes, and that we have a lot to learn about the reasons for this. I'm not sure it's clinical-practice changing at this point. It's more about identifying what the issues are, and I'm hoping it acts as a motivator for future research to look at the factors that might explain the pattern, particularly for Native Hawaiian and Pacific Islander women, because that hasn't been emphasized for ovarian cancer before.

What limitation would you most want to fix?

We didn't know whether patients received guideline-concordant care in the study population. Having information about the treatment they actually received, and whether it met the guidelines for what they should have received, is what was missing. It would be good to have more detailed data on the treatment received in future studies focusing on ovarian cancer survival outcomes.

Has this work changed how you think about your own health, or how you talk with family and friends?

You want everyone to be able to achieve good survival, and race and ethnicity hopefully shouldn't be something that would influence your survival outcomes. I'm always thinking about the fact that not everyone is having the same outcomes. Working in this area has made me more interested in looking for these potential disparities, and in other factors too, like socioeconomic status, or someone living rurally and having to travel a long way to receive treatment. All of these are important factors to consider if you want the best possible outcomes from ovarian cancer treatment.

Ten years from now, what would count as real success for this line of research?

It would be good if clinical guidelines were developed that consider race and ethnicity with more specificity, in terms of special considerations that might be relevant for treatment. And it would be good to advance the biological understanding. There's a lot of work being done looking at patients' tumor tissue to see how the cancer differs in relation to racial and ethnic groups. There's a lot to be done to understand more about the biology as well, to help with better treatment outcomes in the future.

Read the paper

Meagher NS, White KK, Wilkens LR, Anton-Culver H, Bandera EV, Carney ME, et al., Merritt MA. Differences in overall survival for invasive epithelial ovarian cancer by race and ethnicity: results from the Ovarian Cancer Association Consortium. British Journal of Cancer. 2026 Jun. DOI: 10.1038/s41416-026-03509-8

Senior author: Melissa A. Merritt (The Daffodil Centre, The University of Sydney and Cancer Council NSW).

Article citation data retrieved from PubMed.