Over the next several days I became more familiar with both the system and my team. We had six students in total, five ladies and one lad. After watching them shrink into dust on the first day, I saw their personalities take shape as we spent more time together. Mid-way into our second week, they knew to identify topics of interest from their cases for us to discuss together. It’s hard to think that not so long ago I was one of them, trying to stay alive in the face of endless medical knowledge.

After medical school, your knowledge narrows. As a student, you are meant to know a little of everything, with a wide base in medicine, surgery, pediatrics, obstetrics, gynecology, and immunology. Over time as training makes you more specialised, you voyage deeper into your domain, leaving others behind. In America in particular, the internal medicine hospitalist has a fairly narrow range of practice. With the exception of some of the older, career hospitalists that were products of a different era, most of the newer attendings reach very quickly for a specialist consult. This is in part due to the system incentivizing this behaviour; the hospital can bill more when a specialist gets involved, and it also adds prestige to a network to boast highly skilled academics. Here though, the need for internal medicine doctors to be widely trained is clearer. Consult services are thin and strained, and many specialties were new altogether. Nephrology was new for only the past two years, same with endocrine. Pulmonology was less than six months old. Hematology was available in the last two years as well, until the only hematologist left for the private sector, leaving a vacuum. That means for an entire country of 2.6 million people, there were no public sector hematologists. Neurology was also difficult to come by, with only one on staff. Within the first few days, I was getting asked questions about patients that wouldn’t be admitted to medicine at all at Pennsylvania Hospital.

The first was a man in his late forties. He was coming in after experiencing a loss of function of his right arm and some weakness of his right leg. A scan on admission revealed a stroke in the territory of an artery called the middle cerebral, or MCA. This was his second stroke in two years, with one identified previously from an unknown cause. I remember he was a pleasant man and never really had much to say. Every time we rounded on him, he sat up in his bed listening, with one arm flopped over and the other stroking his chin. We had no idea why he was having strokes, given he had none of the typical risk factors of diabetes, smoking, hypertension, or known inflammatory diseases. After doing the basic tests, involving an ultrasound of his carotid arteries which took days to set up, I had almost nothing more to offer when we discussed the case.

‘Here is the reality Neo,’ I remember telling her in response to a question. ‘These patients don’t even come to my service where I work. Even if they did, at this point we would have gotten an MRI brain on him and more sophisticated vessel imaging which I am not even sure we have the ability to do here.’

‘Fair,’ she said.

Another unfortunate reality about strokes is that once the damage is done, it’s hard to reverse the insult unless you act quickly. The often taught notion that ‘time is brain,’ hadn’t caught on here like in the occidental west among the general masses. Patients often presented to the hospital several days after symptoms, meaning that clot busting medications to rapidly dissolve strokes were rarely given, even with large arterial blockages. This gentleman, after spending about 5 days waiting for his ultrasound, was discharged on aspirin and an appointment to follow up in the clinic.

Later in the week we had another patient, a woman in her fifties who came in with the worst headache of her life  and new neck stiffness with vomiting. She was treated at the door as a possible meningitis, with a lumbar puncture and adjacent bloodwork. She also had high blood pressure on admission at 190/110. She was admitted overnight so I only got the chance to see her the next morning.

When we approached her in bed, she was turned over to one side clutching her forehead. As she spoke she squinted, moved slowly to keep her neck straight, and spoke softly. Her neurological exam was completely intact, and she was fully articulate. Otherwise she had no fevers or signs of infection before her symptoms started suddenly. A CT had been done on her, and Neo and I walked the disc down to radiology to get their impression.

‘You see this flaring here,’ the radiologist said to us. ‘This is irritation caused by a bleed in the subarachnoid space. You see it’s not the blood itself, but the irritation that causes the signal abnormality we see on CT. It takes very little to cause this change.’

‘So this is like the hanging chicken sign but not really,’ I said. ‘More like floating chicken wings?’

‘Yes, you could say that,’ he answered.

‘What on earth are you talking about?’ Neo said, turning to me.

‘The hanging chicken?’ I asked. ‘It’s one of the names for the subarachnoid hemorrhage sign on imaging. Let me show you.’

I pulled up a picture from google images and showed her the chicken.

‘I like that,’ she said.

‘Me too,’ echoed the radiologist.

Gotta love a little hanging chicken.

In the end, our patient had a brain bleed, and the question was what to do next. It was hard to say if her high blood pressure on admission was part of the cause, or part of the effect. We got her more sophisticated blood vessel imaging through the private sector that showed she had an aneurysm which was the likely culprit of her bleed. In the end, we consulted the neurosurgical service to review her case and she did fine.

Access to specialists was limited, yes, but this was also exacerbated by poor operational frameworks in the hospital. Calling a consult was a pain in the ass. Sometimes you could get through to somebody and sometimes you couldn’t. Sometimes the specialist took your call, and other times they said fill out a form and we will follow up. Oftentimes you could spend days waiting for a specialist review. Combine this with the high rate of hospital-acquired infections floating around, and you have yourself a recipe for disaster. I had one case like that which left me with a bad taste in my mouth. He was thirty-five, and came in with a swollen hip and a history of being on dialysis.

Dialysis is a fancy word to describe kidney replacement therapy. The function of the kidneys is to clear toxins from the blood through our urine. When the kidneys fail, these toxins build up and can lead to organ dysfunction and metabolic derangements. When kidney failure becomes permanent, the only option outside of a kidney transplant is to put patients on a machine that filters the blood in place of the kidneys. This is called dialysis, and it can be done a few different ways, but most typically involves hemodialysis, where patients come into a dialysis centre three times a week for about five or six hours to get their treatments. Our patient had been on dialysis for not that long until he stopped going due to difficulty ambulating with a swollen and painful hip. He ultimately didn’t seek treatment for his hip or his kidneys, and was finally brought into the hospital when he could no longer walk, after having been off dialysis for ten months. I remember one of the interns was following him, and he presented the case one morning at rounds.

‘This is patient such-and-such presenting with a swollen and tender right hip, and a background history of end-stage kidney disease on dialysis due to uncontrolled lifelong hypertension. Plan for today is going to be X-ray of his hip and consult nephrology to restart dialysis.’

‘Does he need dialysis?’ I asked.

The intern looked at me confused.

‘Yes,’ he said.

‘What for?’ I asked.

‘End stage renal disease,’ he answered. ‘He was on dialysis before and stopped going.’

‘Right,’ I said. ‘And you said it’s been how long?’

‘Uhh… since March so… Ten months?’ he answered.

‘So why isn’t he dead?’ I asked.

‘What do you mean?’ he said.

‘If he has been without dialysis for ten months, and he needs dialysis to survive, then why isn’t he dead? He should be dead right? That’s why patients with end-stage kidney disease need dialysis, is it not?’

‘I guess you’re right,’ he said. ‘So what should we do?’

‘I agree with your plan,’ I said. ‘We need to call nephrology, yes, but the question is not about starting dialysis, our question for him is if he needs to be on dialysis at all. He came in because of his hip. Not because of toxicity from his kidney disease. He still makes urine, his potassium is a little high but it’s okay, and his bloodwork shows some acidosis that maybe we can just treat with oral bicarbonate therapy. So we need to ask them what they want to do for his kidney disease, dialysis or otherwise, and ask them if they will take responsibility for him once he leaves to follow up in the clinic.’

‘Got it doc,’ he said. ‘I understand.’

So we called nephrology and waited. We waited two, three, four days. His hip X-ray was sorted with input from the orthopedic surgeons and several days of follow up and no longer kept him in the hospital. Still, we waited. One week after he was admitted, nephrology finally reviewed him and gave us a plan. His blood electrolytes and acid-base levels, relatively decent on admission, had started to creep up after about the fifth day after he arrived. Still, he was clinically stable, so the nephrologists said they wanted him to get back on dialysis in a non-urgent fashion. Fair, I thought. I was a little miffed that it took a week of waiting, but in the end they gave us a decision which was fine by me. So the medical officer and I spoke to him and agreed to place a special extra wide IV line into his leg called a dialysis catheter. Under my supervision, using landmarks alone, we placed the line on the wards. He was ready for dialysis now and would finally go the next day. He looked like he was doing fine.

Until he wasn’t.

A few hours later after we placed the line, one of the medical students came to find me.

‘Doctor Juan,’ she said. ‘It’s our patient. He is now short of breath.’

Coming to his cubicle, I found the medical officer and two other students at his bedside. Here was our patient, lying with his right hip flexed in the same fashion he had been for a week. Only now, he was heaving his breaths and looked uncomfortable.

Despite his discomfort he was awake and articulate. I felt his pulse and it was regular, not particularly fast. His lungs now had crackles on both sides, most prominently in the right middle section of his lungs. This was new.

‘You saw him this morning,’ I said to the intern. ‘Did he have any pulmonary findings on exam?’

‘I don’t think so, doc,’ answered the intern. He was flustered.

‘Are you sure?’

‘Yes, I’m sure,’ he said.

‘Any urine at all today?’ I asked the group. ‘And does he say this started suddenly or gradually over time?’

They asked our patient in Setswanaa, the local tongue.

‘No urine,’ they said. ‘And it started just now thirty minutes ago.’

‘Any chest pain? Especially when he takes a breath in?’

Again they discussed.

‘No chest pain.’

‘Right,’ I began. ‘Let’s get a pulse oximeter and a blood pressure cuff. I’ll also have one of you get a blood sample for us to check his acid-base status. We need to call nephrology again, this time urgently. No urine now for several days and crackles on exam, this may be fluid backup into his lungs.’

Within short order, we had a set of vitals. His blood pressure was on the low end of normal, around 90/60, and his oxygen saturations were low. We got him connected to an oxygen mask, sat him up, and waited at the bedside for the nephrology team to come see him. His breathing was laboured, but he wasn’t in extremus. I was putting this all down to needing dialysis but there were elements of his story that didn’t add up. His breathing shouldn’t have changed so acutely if he was developing volume overload, and likewise his ankles didn’t look notably swollen, nor did his neck veins. While it was true he had crackles on both sides of his lungs typical of lung fluid build up, the distribution was odd with his right base sounding relatively clear and more crackles higher up. He had also just gotten a central line placed in his leg only hours before, and although we did it with good technique, one of the complications can be an air bubble that travels to the lung and causes effects similar to blood clot. Although the timing of the line was compelling, his heart rate wasn’t high and he wasn’t having the typical chest pain from an embolus. As is the case so many times in medicine, with multiple competing differential diagnoses for his breathing issues, we had to make a calculated decision.

Shortly, the nephrologist on call arrived at the bedside. She was Cuban, like so many of the other imported specialists working at Princess Marina. I discussed the case with her quickly and she agreed, although she had other concerns.

‘We cannot take him to dialysis here with his blood pressure that low,’ she said. ‘We don’t have the ability to monitor him with our staff. I would advise reaching out to the ICU for dialysis under close monitoring with potential blood pressure support.’

When a patient gets put on dialysis, they get hooked up through the line to a big machine. This functionally means that you are expanding the area over which the blood is distributed, like changing containers from a small bottle to a big bucket. Many patients get transient drops in their blood pressure as a result, which is no problem when your blood pressure is 120/80. If you are in the lower range of normal however, you risk bottoming out and having the patient faint, or worse, cardiac arrest.

‘That’s a fair assessment,’ I said to her. ‘It’s unfortunate, his blood pressure had been fine all through his admission until now. We will call the ICU in that case. If they are ready to take the patient, can the machine be brought over to start soon?’

‘Yes of course, the machine is here and ready,’ she said.

‘Very well.’

So we called the ICU. While we waited though, I closely observed our patient. He was starting to fall asleep. I shook him awake several times and he was appearing drowsier. Cycling the blood pressure cuff again, his vital signs were worsening. His blood pressure was now 75/50. He was going into shock. Not to mention his breathing was now slowing, and his oxygen levels were in the low 80s, well below the target of 94-98%.

This is bad.

‘We need the crash cart. Now.’

Neo brought the cart over while I stepped to the head of the bed. One of the nearby nurses helped to plug the oxygen into the bag mask and I placed it over his mouth. With violence I rubbed his sternum and squeezed his trap, he was only flinching in response.

‘Can we get adrenaline ready,’ I said to the nurse.

She nodded

‘And I need you to hurry on getting blood and run to the emergency department to run the gas analysis,’ I said to the student.

‘Yes doc.’

‘Are you looking to intubate?’ asked Neo.

Good question, we probably had good reason to. I for one wasn’t trained to do intubations, medicine residents never are in the US. Here of course though was a different story.

‘I am hopeful that if we normalise his blood pressure his mental status will improve and we can hopefully do fine with oxygen and spare him intubation. If we have to though I am ready for you to take over.’

‘Sounds good, I will get the things,’ she said.

As Neo went, I stayed bagging the patient. He was awake, just very drowsy, and so I timed my breaths to align with his own. His oxygen levels improved slowly. The student who had run the blood to the emergency room came back with a strip of paper and read me the results. Among them, a pH of around 7, meaning worsening acidity of his blood which was not surprising given his worsening renal function and acidosis from toxins. What I didn’t know was whether or not he was producing lactic acid, which is a common occurrence in sepsis. The lab here couldn’t run for lactic acid at all. Could this be sepsis?

‘I think his acidosis is likely not helping the situation, so let’s give a few vials of bicarbonate, have you got them in the cart?’ I said, turning to the nurse.

‘Yes,’ he answered. ‘How many do you want?’

‘Let’s do four,’ I said. They were small, each with about 20mEq of bicarbonate solution.

As we bagged him, slowly he began breathing more rapidly. I still bagged him with each breath, but I saw some life return to his eyes. The medical officer asked him questions and he answered. Checking his blood pressure again, it was now closer to normal. This confirmed our suspicion that his drowsiness was mainly attributed to his shock state and low brain perfusion, and not some other metabolic derangement. Yes he was looking better for now, but that was with 80 units of bicarbonate and adrenaline in his system. He needed urgent transfer to the ICU now more than ever.

I stayed at his bedside, in plain view of all the other beds in the cubicle, recycling his blood pressure and making sure he was awake for close to an hour before the ICU team arrived. It was a tall Cuban doctor, with thin features and wire glasses, and a young local surgical resident with a large personality. I had interacted with them before in the hospital and they knew what I was about. I explained the case to them and they agreed with everything. Yes, needs dialysis, yes, should be in the ICU, yes, not needing intubation right now. However, we had no room in the ICU at Princess Marina, so he would have to be outsourced. Neo stepped aside to call the nearby private hospital for admission, this was the regular practice here when our wards were full. The government would then foot the bill for private care.

‘Let’s look at the chart,’ the Cuban doctor said.

They both flipped through paperwork.

‘Look at this,’ the resident said. She was pointing down.

The last three blood draws recorded, over the period of the last three days, each showed a gradually rising white blood cell count. Today’s result was the highest of all.

‘Sepsis,’ the Cuban doctor said.

He was right. The story made more sense this way. In all likelihood, a hospital acquired pneumonia had been brewing for some time. It was likely the reason for his worsening kidney function over the last few days as well given he went almost a year without dialysis. It would also explain his shock state, acutely worsened acidosis, and his chest exam. How did I miss this? I thought to myself.

The answer though was simple. I missed it because it was easily missed in a system full of room for error. A mess of loose papers in a chart that goes missing for hours at a time, an intern covering a patient out of his depth, rounds that drag on endlessly, and a one dimensional admission in which all we did was wait for days meant people stopped paying attention. When I asked the intern about the rising white cells, he had no idea. If we had known, we could have acted sooner, evaluated closely for infection or started antibiotics earlier at the first signs of trouble. At least we could have had a plan in place for a potential decompensation. Instead, here we were, day seven of an admission that should have lasted three, and our patient was in septic shock being urgently transferred to another facility. It was an encapsulation of all the tiny frustrations of Princess Marina Hospital that felt highly preventable.

It’s not surprising he got a hospital acquired infection; they were so common in this hospital it wasn’t even funny. Most of my exams were done barehanded as we had no gloves. Many of the sanitizer dispensers were out intermittently, and patients sit within meters of each other, visited by family four times a day. It’s a bacteria’s paradise. Unfortunately, I would find out the next week that our patient died of sepsis a few days later at an outside ICU.

His case really bothered me. It bothered me in ways that working in worse settings didn’t, and I tried to figure out why. I think it’s because Princess Marina has the chance to be a great hospital, but it gets in the way of itself at every opportunity. The clinical quality of the healthcare staff is truly second to none, and I mean that. It’s at the intellectual level of the highest academic institutions both in the US or Ireland. Morning conference is a smorgasbord of diverse academic opinion and thoughtful discussion on complex cases. The doctors care, a lot, and work their asses off. And yet, why do patients like this guy die when they shouldn’t? The answer, I think, is in medical infrastructure and poor auxiliary services. When your technology is down half the time, when you don’t have a standardised way that results are reported and followed up on, when you don’t even have a proper bookshelf with numbered spots for patient charts, when your vital signs aren’t documented in a way that’s intuitive, when your consult services don’t have a pager, when you don’t have gloves, when nursing doesn’t give medication because of one reason or another and you don’t find out until several days later by chance when you ask why there are doses missing, when teams don’t have clearly defined roles or a system for tracking jobs, when people don’t even write legibly in charts – the list is endless. The problem is not lack of expertise, the problem is lack of coherent frameworks. More funding would help, but I don’t think that’s the crux of the issue. I had little to offer Princess Marina Internal Medicine. As a matter of fact, I had far more to learn. What Princess Marina needs is help refining their systems.

This made me realise that there are different levels of healthcare system development, and the playing field is not always level. In Uganda, at a peripheral site with an overwhelmed medical officer and overworked staff, it seemed inevitable that patients have bad outcomes, but here it seemed more unforgivable. The patients deserved better. Somewhere like a refugee camp in Uganda, a good framework is of little use to physicians who don’t have capacity to provide quality care. Somewhere like Gaborone, where their education system has clearly produced fantastic doctors, it feels like all of that expertise gets nullified by systemic errors that worsen outcomes. Determining where a health network is on that spectrum is what should inform areas for development and capacity building.

The death of our patient spurred me to ask more questions of Princess Marina. I scrutinized everything more closely, although my comments were internal. I would keep them sealed until the right place and time for an audience with the right people. I did, however, press the team more to make progress where I could. I had this case as reference for every time we floundered.

‘This is why we need to move people out of here,’ I’d say. ‘The longer they stay the bigger the risk.’

So maybe the answer for Princes Marina is two-fold. More specialists is one, definitely, as after all, if there were more nephrologists, our patient likely would have been seen sooner, or never been admitted in the first place with closer follow up in the clinic. The other is more refined healthcare infrastructure. This is something that I am not an expert in, and is much easier said than done. Some things seem easy, and others multi-faceted. What I can say though, is that we need to give them a chance. Despite all my critiques, we are talking about a nation that was born in 1966. The US and Europe had already built, rebuilt, and conquered the world ten times over by then. The hospital I work in was built by Benjamin Franklin and Dr Thomas Bond in 1751 for Christ’s sake. Yes, their services are lacking, yes, there are flaws, but also, they’ve come a long way, let’s give them a chance. Offer help, and give them a chance.

That’s the least we can do.