If you are divorced, widowed or never married and develop cancer, watch out. You may get less aggressive treatment than your married friends.
We've often heard about studies showing that married adults are more likely to survive cancer than singles. But buried in those same studies is another finding that hasn't made the headlines. When surgery or radiotherapy is the treatment of choice, patients with spouses are more likely to get it.
I had no idea that marital status might affect medical care until an oncologist, talking about what treatment to give me, asked if I have a spouse or children. When I said no to both, he looked genuinely concerned. "But how will you manage?" he asked. He then proposed to give me only one mild drug, although the standard of care was a much harsher — and more effective — combination chemotherapy. When I tried to describe my strong network of friends and extended family, he talked right over me.
If I hadn't moved on to another oncologist who gave me the recommended treatment, I probably wouldn't have survived.
As an experienced researcher, I was curious. Was this just a freaky personal experience, or are single patients often treated less aggressively?
To learn what cancer experts have to say about that, I reviewed 59 studies based on the Surveillance, Epidemiology, and End Results Program (SEER), a massive database maintained by the National Cancer Institute. Cumulatively, these studies cover 7,331,695 patients with 28 kinds of cancer.
With a background in literature, psychology and law, I wasn't about to comment on oncological findings or statistical methodology. But when study after study reported significant differences in treatment rates between married and unmarried patients, I did want to know how the authors explain that discrepancy.
One proposed explanation is that patients who don't have surgery or radiotherapy must have refused it. So authors describe patients as "avoiding" treatment or as not "accepting," "pursuing" or "submitting to" it. Unmarried patients may lack "the fighting spirit," they suggest, or perhaps "having a committed partner gives patients more to live for."
But that explanation doesn't stand up to scrutiny.
Using SEER data on 925,127 patients, researchers from Harvard, MD Anderson Cancer Center and the Mayo Clinic, among others, found that only 0.4 percent declined surgery when physicians recommended it, and 0.9 percent declined radiation. Unmarried patients were indeed more likely to refuse, but the proportion was small. Of 278,015 unmarried patients whose physicians recommended surgery, 1,441 refused. For radiation, it was 1,055 out of 79,303.
Conspicuously absent from these studies is any analysis of the physician's role in recommending treatment. And yet, the authors' speculations about why unmarried patients are systematically undertreated bear the unmistakable imprint of widespread stereotypes.
Consider, for instance, an often-cited study in the eminent Journal of Clinical Oncology, reporting on 734,889 patients with 10 kinds of cancer. Like other such studies, it associates unmarried status with depression, social isolation and noncompliance with medical instructions, with no disclaimer that that profile might not fit all 110 million unmarried adults in the United States, out of 252 million over age 18.
Since SEER provides no information on psychological characteristics, I checked the footnotes in the medical studies to see what data the authors were using to support the contention that unmarried patients are systematically undertreated because of their own vulnerabilities.
Here is one example, from the JCO article. Like other studies, it portrays unmarried patients as so prone to depression that perhaps they should be routinely screened for it. The source cited to support that assertion is a 1996 comparison of depression rates among nations. Its only relevance to marital status is the finding that 4.3 percent of U.S. spouses are diagnosed with major depression, compared with 11 percent of divorced/separated adults.
The JCO authors did not acknowledge the absence of data on widowed and never married people, or that most divorced/separated adults are not depressed.
Research also casts doubt on the conflation of marriage with social connectedness. Of course, high-quality marriages do provide significant emotional and practical help. But social support comes from many sources, and studies show that binaries such as married/unmarried are the least effective way to measure the association between social support and health outcomes.
But the authors go even further, speculating that "[I]t is possible that unmarried patients are innately different from married patients. . . . [And] caution should be exerted before assuming that improved social support would significantly improve outcomes in unmarried patients."
Prominent among sociologists who agree with them is Linda Waite, who co-wrote "The Case for Marriage: Why Married People Are Happier, Healthier, and Better off Financially."
"In the U.S., where people have plenty of options for marriage, it's likely to be those who are disabled or otherwise at a disadvantage who don't marry," she said. "And so, they might indeed do worse in health care because of the underlying issues that caused them not to marry." In her view, medical providers are justified in giving significant weight to marital status, "because unmarried patients are unlikely to do well, and they want to take patients who will do well."
Many sociologists say otherwise, however.
As a meta-analysis of 148 studies of social support and survival concludes, "while researchers may be tempted to use a simple single-item such as 'living alone' as a proxy for social isolation, it is possible for one to live alone but have a large supportive social network." A study based on the National Social Life, Health, and Aging Project bears out that result, finding no difference in depression, loneliness or psychological well-being between partnered and unpartnered older women.
Interestingly, at least some authors of SEER-based studies recognize as clinicians what they ignore as researchers. Ayal Aizer, a Harvard radiologist who led the JCO study, was quick to say that some patients "don't have family but have supportive friend networks," and physicians should ask about nontraditional support before determining treatment. And, he acknowledges, it's indeed possible that time-crunched physicians might substitute marital status for a broader assessment of social support.
Nevertheless, he doesn't see a need to examine how often physicians recommend surgery or radiotherapy to married vs. unmarried patients. Even if such research revealed discrepancies that don't seem justified by factors such as age, health and cancer status, he said: "You never know what goes into a decision. It could be justified."
He's right, of course. There's a limit to what can be proven simply by quantifying how often physicians recommend treatments to different groups, although the results might inspire more in-depth research. And even with unbiased treatment, unmarried patients might remain at a statistical disadvantage because of factors such as insurance, finances, and the unmet needs of those who really are isolated.
Still, the overgeneralized, unmitigatedly negative portrayal of unmarried adults is cause for concern, particularly when based on nonmedical research that is outdated or only tangentially relevant. It is, in effect, an invitation to associate unmarried status with a need for milder treatment, without doing enough to ascertain whether an unmarried individual could handle a more aggressive approach.
So I asked experts in sociology and psychology: How might the best work in those fields contribute to medical research?
Psychiatrist Jonathan Metzl, author of "Prozac on the Couch," says doctors see stereotyping "as bad, something we're supposed to eliminate." Realistically, however, forming shorthand judgments is "inherent in human interaction, and believe it or not, doctors are human too." The most professional approach, he maintains, is to "Get it out on the table, talk about what's influencing a particular decision. It's good to be self-aware of one's own assumptions."
"Frame the discussion in terms of what the patient actually needs, rather than focusing on whether it's provided by people in specific roles," says Susan Brown, co-director of the National Center for Family & Marriage Research. "Our whole system is built around traditional family roles, and that doesn't work for many people."
Metzl agrees. "The more social support, the better. But the minute you link that to marriage, it conveys a series of social assumptions."
Leslie Hinyard, deputy director of the St. Louis University Center for Health Outcomes Research, describes herself as "a huge proponent of the interdisciplinary team approach." The ideal study, she says, would combine physicians' expertise in oncology with a social scientist's broader understanding of the possible reasons for discrepancies based on marital status.