Being a doctor is the only thing I ever really wanted to do. I remember watching Dr. Kildare on TV as a kid and saying, “That’s going to be me.” I started practicing in 1967 and fell into my work completely. Medicine became my life. I was what you might call an old-time physician: I got to know my patients personally, I studied their histories, I preferred physical diagnosis to scans and ultrasounds. Though it had fallen out of fashion, I loved making house calls. I loved rushing through the city, syringe in hand. I became involved in the lives of my patients and sometimes cared for two or three generations of their families. I saw them through difficult times. When they were sad, I was sad. I cried with them. I had some prominent patients over the years, too: Studs Terkel, Jesse Jackson, Carol Moseley Braun. I was Barack Obama’s personal physician for 20 years. My practice was my sustenance.

I retired on September 1, 2014. Ten days later, my wife, Margaret, was diagnosed with pancreatic cancer. She was 71.

She had suffered a spinal injury a few months earlier and had been in physical therapy. One day her therapist remarked that she looked jaundiced. I’d been sitting across the breakfast table from Margaret every day and hadn’t even noticed. I took her to Loyola University Medical Center, where doctors, using ultrasound and then a CT scanner, detected a mass in the pancreas. Margaret was sent home after doctors performed a biopsy, and a short while later the results confirmed the worst: The mass was cancerous. Margaret and I were both deeply shocked by the news, and I was terrified at the prospect of losing her, and of her being in pain. I was angry, too—angry at God for putting me through the same thing twice. I’d lost my first wife, Polly, to colon cancer in 1993. 

This time would be different, though. Since I was no longer practicing, I would be able to turn my full attention to Margaret—not only as a husband but as a doctor. For the next year and a half, in an ordeal that would test both of us in every possible way, I would serve as her primary physician. 

At the very end of my career, at the age of 77, I suddenly had to face the most daunting medical and emotional challenge of my life. The experience taught me a lot of things about myself and also about medicine in general—namely, what a difference continuity of care and close observation can make to people with a severe chronic illness. For me, it also shed light on a medical system that too often treats patients more as cases than as people. 

Margaret had platinum-silver hair and was an exquisite beauty. I’d asked her to marry me just six months after meeting her in 1994, but she made me wait four years because she wasn’t convinced I’d gotten over the death of my first wife. As a psychologist (she was a pioneer in the treatment of autism in children), Margaret had more insight into these things than I did. She was tough, too, and always faced life’s obstacles with grit and courage, which made it all the harder to see her weaken so quickly. 

A few days after getting the biopsy results, before we’d even figured out a treatment plan, Margaret started vomiting. I rushed her back to Loyola. It turned out the tumor was obstructing the first portion of the duodenum, which is the part of the small intestine that comes right after the stomach. The mass was effectively blocking everything from going through. 

Doctors put in a stent to keep the bile duct open, and that at least got rid of the jaundice. But unable to take any foods or fluids by mouth, Margaret quickly grew weak from lack of nutrition. She underwent surgery to alleviate the blockage and allow a feeding tube to be installed, but the procedure also revealed that the cancer had spread to the peritoneal cavity surrounding the abdominal organs.

It was devastating news. I didn’t think she would live past Christmas. 

During those excruciating early weeks in the hospital, Margaret’s doctors consulted closely with me, showing me her scans and bloodwork. But what I found startling was how rushed they were. They were clearly good physicians, but they were spread too thin, with too many patients to see, and seemed handcuffed by a system that prevented them from always acting on their best instincts. They didn’t have time to give my wife a full physical exam. They were quick to rely on scans and other diagnostic tools and seemed uninterested in physically touching their patient. They didn’t ask me for her full medical history. (I had to volunteer the information about her spinal injury because I knew that could affect treatment.) No one checked her lymph nodes. And inadequate attention was paid to Margaret’s electrolyte balance, particularly her levels of potassium, one of the most important chemicals in the blood. I knew that the potassium being lost through the suction in her stomach had to be not just replaced but maintained on a daily basis. My encounters with the attending physicians were tense. I realized I was infringing on their turf, but I also felt I had no choice. 

Margaret’s experience during those first weeks in the hospital confirmed for me something I already knew: Continuity of care is becoming a thing of the past, at least as far as severe illness is concerned. Increasingly, when you go into the hospital, your care is handed off from your usual doctor to a rotation of attending physicians, or so-called hospitalists. It’s not unusual to have, as Margaret did, a sequence of different doctors. Technically, most of them are good. They know what to do when a lab test is abnormal. They know how to perform exams focused on the patient’s problem area. But they often don’t have a sense of their patient as a full person. Hospital patients are human beings at their most vulnerable. And so they want a doctor who knows them, who understands their medical history, their personality, even their family dynamics. That sense of connectedness is getting lost. 

But not entirely. After Margaret was transferred to a long-term acute care facility at Presence Holy Family Medical Center in Des Plaines, where she would spend more than two months, she came under the care of an oncologist named Sasikala Paidi. An Indian-born specialist in her early 50s, Dr. Paidi was attentive and kind. She would sometimes spend an hour at Margaret’s bedside, teaching her ways to alleviate pain and discomfort. She counseled my wife emotionally, even spiritually. She counseled me, too. This was a physician who seemed to share my belief that cancer is a systemic disease and requires a systemic approach. 

Margaret did live past Christmas. And in fact, by February 2015, more than five months after her diagnosis, she had recovered enough strength that the doctors said she could leave Holy Family—with the understanding that I would take charge of her day-to-day medical care. In that regard, Margaret was fortunate to be married to a doctor. The kind of observation and intervention she would require was more than any nursing home—let alone home health aide—could handle: 24-hour feeding pump, intravenous line, various injections, electrolyte and fluid balance checks, blood draws, multiple medications, constant monitoring for fever and infection. I’d installed a hospital bed, complete with IV poles, in our family room, which had a TV, lots of plants, and plenty of natural light.

The long-term-care insurance we’d bought a while back paid for a nursing aide for 12 hours a day; she came from 7 a.m. to 7 p.m. I took care of Margaret for the other 12 hours, with the help of Margaret’s two daughters, Valerie and Michelle. But I was on call 24/7. I even placed a baby monitor next to Margaret’s bed so that I could be at her side at a moment’s notice.

Things began to go wrong quickly and with numbing frequency. During the months to come—and I say this not to boast but to give a sense of cancer’s volatile unpredictability—I would literally save my wife’s life multiple times. 

Margaret had lost a lot of weight and was receiving fluids through a long-term catheter called a PICC line, which is threaded into a vein just above the heart. She was prescribed heparin to keep blood clots from blocking the PICC line, but Margaret reacted badly to the drug and had to go off it. To keep the line clear, I would flush it with a saline solution a couple of times a day, but still the line clogged. This was an urgent situation because it prevented her from getting any intravenous fluids. Dr. Paidi would prescribe something to dissolve the clot and we’d be OK for another three weeks or so, but then it would happen again. I became fixated on watching the PICC line. This caused enormous stress.

So did problems with Margaret’s other lifeline: the feeding tube. Inserted through the left side of her abdomen, it was how medicine was administered. The tube wasn’t big enough to accommodate whole pills, so they needed to be ground up. But I’d learned in the hospital that even ground-up pills could clog the tube, requiring its removal, which in turn required a surgical procedure. I’d begged the doctors not to administer pills through the tube, but often this request would not get passed on from one attending physician to the next. Now, at home, I consulted the formularies of pharmaceutical companies in order to find Margaret’s medicines in liquid form. 

Late that winter, Dr. Paidi started my wife on a single-drug chemotherapy. She wasn’t strong enough to take a full-court press. After a time, blood tests and periodic CT and PET scans showed that the tumor didn’t seem to be growing and was possibly even getting smaller. We added another chemo drug. It made Margaret anemic, so she had to start receiving injections to build her red blood cell count. Her white blood cell count was getting knocked down, too, so I began injecting another drug to deal with that. We did blood counts twice a week, which required special care because the blood had to be drawn from the PICC line. And since the lab seldom contacted me with the results, even though the information was essential to Margaret’s care, I had to stay on top of calling them. 

Margaret quickly lost all of her beautiful silver hair and became so thin I could easily wrap my fingers around her arm with room to spare. She developed bedsores. But she never once griped about these things. Her only complaint was the pain. We began to manage it with a patch that released fentanyl, an extremely strong drug. Over time her discomfort required higher and higher doses, to the point where she reached the maximum-strength patch. It still wasn’t enough.

Dr. Paidi then prescribed a fentanyl spray administered under the tongue. In keeping with security protocols, the highly addictive drug had to be delivered directly by a representative of the drug company. When the rep told me the retail price—about $8,000 a month—I blanched. Our insurance would pay only a portion. But the rep put me in touch with Patient Services, a donor-funded nonprofit that provides financial assistance to the families of chronically ill patients. The organization picked up the remainder of the cost. I felt lucky. Many caregivers probably never realize they have this option. 

Margaret’s escalating pain, and the increasing doses of drugs needed to combat it, became a constant source of worry, but other crises distracted me. At one point Margaret began to run very high fevers, which meant trips to the ER, where we’d wait for sometimes five or six hours to have blood cultures taken. It turned out the fevers were being caused by sepsis and had to be treated with an IV antibiotic—yet another thing to add to our daily routine. 

The fevers soon returned, so an infectious-disease specialist and I, under the assumption that Margaret’s body had built up resistance, tried a different antibiotic. The antibiotics sometimes gave Margaret diarrhea, and I’d have to come downstairs during the night to clean her. To complicate matters, the second antibiotic was not covered by our insurance, and we had to pay thousands of dollars for it. Unlike many people in similar situations, we could afford it, but this added to our stress nonetheless. Margaret would remain on IV antibiotics every day for a year and a half—I’d never seen anything like it in all my years of practicing medicine. 

About two months into the chemo, another problem developed: restless legs syndrome. It was the most malignant form of this neurological disorder I’d ever witnessed. My wife’s legs would thrash madly. She’d be in agony and force herself out of bed in the middle of the night just so she could move her limbs. I researched treatments frantically, calling a neurologist I knew, and eventually put Margaret on a movement disorder drug normally used for Parkinson’s disease. When things got really bad, I’d resort to giving her intravenous Valium, which I had to obtain directly from the hospital, since pharmacies don’t carry it. In time, we got the restless legs syndrome under control. It felt like a triumph.

The cascade of crises wore us down. Margaret, strong as she was, had become so fragile that I was always on alert for the next emergency. In between dealing with them, I had to constantly get help from Michelle, who is a software engineer, with the settings of the electronic feeding pump, and I had to monitor Margaret’s electrolytes and study the results of her kidney function tests to make sure she was getting enough intravenous fluids. Then, to make sure I wasn’t overdoing it, I had to remember to measure her legs to see if there was any swelling. Even as a first-year resident, I’d never had to deal with this kind of workload, and never in my career had I faced so many complex and seemingly unsolvable problems at once. 

I look back now and am amazed at some of the decisions I was able to make under such difficult circumstances. I also often wonder what Margaret and I would have done if I hadn’t been a trained physician. Our Medicare would have soon stopped paying for Margaret’s stay in the long-term acute care facility. Where would she have gone? No nursing home could have provided the kind of around-the-clock monitoring and crisis response that she needed. And even when Margaret was home, I literally had to count the days before we could readmit her to a hospital, because our in-patient coverage had been maxed out and would only kick in again after a 60-day hiatus. And what would we have done if we hadn’t opted for the supplemental insurance that paid the $300-a-day cost of a home health aide? 

The fact is, even with the best care and the best insurance, there’s not a system in place in this country to properly manage the type of long-term illness my wife had. 

The median survival rate for pancreatic cancer is eight months. Margaret passed that milestone in May 2015. Then 12 months. Then 18. Those months were hard, but they gave us more time together, more time for her to spend with her children. It gave her time to learn she was going to have a granddaughter, who would be born just weeks after Margaret died.

There were even moments of relative normalcy. When Margaret felt strong enough, she would walk upstairs to her office to write emails, with me or our aide climbing the steps alongside her, carrying the IV poles. Once in a while we’d switch off the feeding pump and the monitors and go outside. Valerie, who would visit two or three times a week from her home nearby, got Margaret a special backpack so we could carry her feeding pump and stay out longer, taking strolls, with the help of a rolling walker, in our Northwest Side neighborhood and on the 606 trail. That summer of 2015, Valerie even bought her a drysuit so she could go in our pool. My wife loved to shop—nearly as much as I hated it—and sometimes I would take her to stores just so she could browse the aisles. It gave her a second wind, and sometimes we’d shop for hours. 

Occasionally, when her pain wasn’t too bad, I’d get in bed with her just to lie beside her and touch her, have my skin in contact with hers. I probably spent more time with her during her illness than in the whole 15 years of our marriage before that. With the exception of picking up prescriptions or buying groceries for myself—Margaret couldn’t consume anything by mouth except for an occasional taste of juice or ice cream, which would just come out the drainage tube after—I almost never left her side.

We still argued about small things, like we always had. She was meticulous about her possessions and her papers and yelled at me if I moved something without telling her or threw away old bills or letters. But eventually even that started to change. 

Early in the spring of 2016, blood started coming out of the drainage tube in her stomach. First it was just a little, but soon the volume became significant. I took her to Advocate Lutheran General Hospital, and doctors there discovered huge varicose veins in her stomach. They performed an extraordinarily complicated procedure that involved threading a catheter into the liver and ultimately into the portal vein—an operation I’d never seen performed before—in order to put in a stent. Miraculously, they succeeded in stopping the bleeding. 

But the operation took a toll on Margaret. In the days after the procedure, she grew weaker and weaker. It became obvious that things were getting bad. After about 10 days, my wife made a decision: She wanted to stop the feeding, stop the intravenous fluids, stop it all. 

I prayed. I’d never been a very good Jew—I went to services only on the High Holidays—but since Margaret’s diagnosis I’d started praying almost every day to ask God to get us through the various crises. But now I was asking for something different: I was praying that she would die. She had survived 20 months, but now life had become agony for her. And it was agony watching her in pain.

We transferred Margaret to a hospice, where she hung on for more than a week, growing weaker and less aware with every passing day. Friends would say, “She has such a love of life, she’s clinging to it.” I didn’t believe that. Her hanging on was just another cruel part of her illness. 

At some point near the end, I remember, I told her again that I loved her.

She said, “I love you too.” 

Those were her last words to me. 

I was with her a couple of days later, on May 4, 2016, when she took her last breath. I called the funeral director, then helped put her in the body bag. It just felt like the right thing to do, my last duty as her doctor.