Reflections on the analogs between human and environmental health.
A couple of weeks ago, my 83-year old dad, who is staying with me, complained of extreme weakness. I took him to our family doctor who ran a bunch of tests; blood work, X-Rays, ECG. The X-Ray showed “bilateral pleural effusions” in the lungs, characteristic of Covid-19. We rushed to the Covid emergency room (ER) with a letter from our family doctor. The next 12 hours in the Covid ER were harrowing, but that is a story for another day.
The first antigen test was negative. So we had to wait till the more reliable RT-PCR results came out the next day. At midnight, the ER staff decided not to wait, but to admit him into a Covid ward to start anti-viral treatment, given his age and co-morbidities. However, (to give the hospital credit) they were careful to give him a private room and not enter the room and expose him to the Covid medical staff, just in case he didn’t have it.
The RT-PCR test was negative the next morning. Suddenly, the circumstances flipped. The hospital immediately moved him to a non-Covid room. The problem is, they were still certain that it was a false negative. So for the 5 days (till they could repeat the RT-PCR and ensure he wasn’t exposed in the ER), everyone, hospital staff included, continued to pretend he had Covid, while also treating him as if he did not. Since this is pretty much how we seem to lead our entire lives these days, I realized how apt the term Schrodinger’s virus (not my invention) was. We all must behave as if we all have Covid to stop infecting others. Yet we must remain calm in the belief we do not have it, to avoid rushing to the hospital with every sneeze!
But this blog is not really about Covid. While all this was going on, I was simultaneously working on an article and organizing a workshop on Bangalore’s lakes. There were so many parallels in the conversations that I thought it worth documenting.
Diagnosing the problem.
So when the doctors first looked at the blood tests, they showed low hemoglobin and high levels of infection markers. The X-Ray showed bilateral pleural effusions (that other forms of pneumonia also display), but we are in the middle of a pandemic so what the odds favored Covid-19 vs. other virus strains. We also consulted friends in other parts of the world who offered rules of thumb like “if the neutrophil level is high, bacterial pneumonia is more likely, if only the lymphocyte count is low, Covid is a possibility”.
The ER doctors were very young (in their 20s) and exhausted. I wondered whether they had the skill to look at tests and distinguish between viral and bacterial pneumonia (they were indeed confused) if they had not seen enough cases of other types of pneumonia in their lives. Moreover, although there are millions of Covid case histories around the world, the data aren’t collated into clearly defined probabilistic diagnosis pathways. As I probed, I realised there isn’t any method to it, rather a fuzzy-set of evolving best treatment practices that are somewhat inefficiently communicated and shared.
While I was waiting for the Covid test results, a researcher on my team, sent me the following maps of nutrient levels in Jakkur Lake. Jakkur gets most of its water from a 15 MLD Sewage Treatment Plant (STP); the treated wastewater enters the lake via a constructed wetland. During heavy rains, it also gets water from two inlet stormwater drains that occasionally experience raw sewage leaks. I could not understand why the nitrates and phosphates were showing contrasting spatial patterns if the source of both nutrients was the same. Was nitrate getting resuspended from the sediments? Was the lake safe from harmful algal blooms, because it was P-limited?
Not being a water quality expert, I asked my colleague Priyanka Jamwal, who responded that while both nitrates are phosphates originated in the STP effluent, the phosphates get attached to suspended solids and removed by sedimentation. In contrast, ammonia and other N based organic compounds get converted into nitrates as water moves from inlet to outlet and remain in the water. Importantly, BOTH nitrates and phosphate levels were too high. The first step was therefore to lower the nutrient levels overall, and then see which nutrient is limiting and can be controlled further. Finally, although the phosphate levels were lower towards the outlets, even the 0.9 mg/l P levels reported at the outlet was too high. To ultimately limit the growth of harmful cyanobacterial blooms, it would need to be below 0.1 mg/l.
Designing an intervention strategy.
Based on the tests, the doctors decided to proceed with both anti-viral and anti-biotic treatments. They also decided to address his anemia through Vitamin B-12 and iron injections. Finally, they put him on oxygen to address his low blood oxygen levels. In deciding the intervention they also considered his age and other factors to ensure his body could handle the prescribed treatment and his quality of life goals.
Our attempts at designing an intervention strategy to address the high nutrients for the lake were less successful. Three discussions clarified for me how little our collective understanding of the efficacy of interventions is. First, a fierce debate erupted in one of the citizen groups, whether water hyacinth should be introduced into lakes to remove nutrients. Some ecologists were of the opinion that it is so invasive that it ought not to be attempted. Others felt it would be more effective in scrubbing nutrients and would help prevent harmful algal blooms. The question remained unsettled even after much debate. Second, we had a call with one company that designs floating wetlands. They argued that most of the organic matter and nutrient break-down occurs through bacteria in the matrix of the floating wetland and not the plants themselves. They were able to share peer-reviewed research on the efficacy of floating wetlands. The problem was there were no design specifications to determine the size for a 50-acre lake, with given nutrient levels or which plant species are most effective under different conditions, or predict what decreases might be expected. They were marketing the wetlands primarily on aesthetics. Third, unlike medicine, my colleague Priyanka argued in an email, the process of setting goals for lake restoration with stakeholder inclusion is completely absent. The absence of goals and design protocols leaves much room for uncertainty, distrust, and in-fighting.
Alternative herbal treatments
So frightening was the Covid ER experience, that the first thing I did when I got home was to try and boost my own immunity. My friend Mr. Ramprasad (founder of Friends of Lakes) was kind enough to send me an ayurvedic “kasayam” (herbal tonic) and I started “jal neti” (nasal irrigation) right away. Do these treatments work? There is no scientific evidence supporting nasal irrigation for Covid, though it may help with common colds. But I felt it fell into the “do no harm” category, with clear side benefits, so I rolled with it.
A few days prior, I had had a call with the founder of Cownomics. Among other treatments, they use a traditional mixture of herbs to treat wastewater in ponds and lakes. They claim that these have been extremely effective. Their basic premise is that bacteria present in these, are effective at breaking down sewage. Given how bacterial starters have revolutionalized home composting in Bangalore, it seemed plausible. The problem is they only had limited data on efficacy, done in small settings. In the absence of third-party field tests in an RCT setting, I could not see a way forward for a large scale implementation (they are working toward this).
Treating the cause vs. treating the symptoms
The third RT-PCR test was also negative. So we decided to bring my dad back home and continue the antibiotic treatment. The problem is he was unable to sustain his blood oxygen level much above 90 and my dad was stressed about it. We arranged for a home oxygen concentrator, but thanks to terrible communication on part of the rental agency, we did not actually connect the oxygen correctly the first night. We realized the next morning he had just been breathing air. After long arguments with the dealer and frantic calls to the doctor, we collectively decided that he did not need the supplemental oxygen at all. We decided not to chase the magic number 95, given his specific circumstances and comfort level (though we have a working oxygen concentrator borrowed from a family member, just in case).
Earlier in the week, our colleagues at Biome Environmental Solutions had described their experiments with solar aerators at Jakkur Lake. The aerators did seem to elevate the minimum Dissolved Oxygen above hypoxic levels, but don’t of course address the underlying problem. It is also worth noting that Jakkur Lake is generally considered a success story, hosting a large bird population and a vibrant lake community, notwithstanding the presence of occasional dead zones. While the data on the efficacy of aerators is not out yet, perhaps we need more subjective measures of lake health too, instead of focusing on 4 mg/l at all times and locations.
The role of expertise and rights of stakeholders
I got into a bit of a spat with the doctors on Day 3 of the hospital stay. They had ordered a bunch of tests (precipitating a minor emergency), with no information to any of us. I argued that they needed to be cognizant of the patient’s right to consent for non-emergency tests. The hospital was careful to keep us informed after this conversation, but it was clear that “doctors know best” is the way hospitals typically operate.
That same day, one of our collaborators called me fuming that government engineers had formulated a ridiculous project plan for one lake (the drawing essentially expected the water to flow upstream in one place) but were refusing to let citizen groups question their expertise. In the same call (with no irony at all), while we were discussing whether a closed-door session on lakes was appropriate, the same collaborator expressed his frustration with large public sessions where “someone who has only worked on lakes for a day has the same amount of floor time as someone who has worked on lake rejuvenation for years”. The question was who has authority on technical matters and who can call the assumptions into question.
Unlike a doctor-patient relationship, where there is only one stakeholder (the patient), in a democracy, everyone has the right to participate in public discussions on public commons, even if they are uninformed and disruptive. But this frequently creates stalemates with citizen groups suing and counter-suing each other. What we urgently need is a communication language, guidelines, and structure for engagement. If there are established best practice manuals, people can then question agencies on the deviations from them. But at least we won’t have each group reinventing engineering standards from scratch.
The case for implementation science and a culture of “learning”.
It was fascinating to me, how much the doctors view the patient as a “data stream” rather than a living being with rights and agency. Further doctors rely primarily on personal expertise to wade through the diagnosis process, calling on more experienced practitioners when they get stuck. But the entire process depends entirely on the data assimilation and application of logic by individual humans.
Perhaps one day, AI will take this role over, collapsing the entirety of medical knowledge into probabilistically well-defined decision trees. Maybe then the doctor can be liberated from crouching over reports to really seeing and communicating with the patient. But even without AI, in his excellent book “The Checklist Manifesto”, Atul Gawande from the Stanford School of Medicine shows how simple checklists can solve the doctors’ diagnostic problems, provided the check-lists are carefully curated to reflect the state-of-the-art medical knowledge and the collective wisdom of doctors’ hospital experiences.
In the water sector though, we are years and years behind. We barely have any empirical data let alone something resembling an evidence base (RCT or otherwise). We have not synthesized the existing sparse evidence into decision trees. Our engineers are not trained to communicate with stakeholders, justify decisions, or indeed ask for what their restoration goals are. The furor over the appointment of NEERI (a scientific organization) to come up with a master plan for Bangalore’s lakes just today, is a case in point. And then we implement projects but rarely track their effectiveness.
This problem was addressed in the medical field years ago, by creating “Implementation Science”, the scientific study of methods and strategies that facilitate the uptake of evidence-based practice and research into regular use by practitioners and policymakers. It was introduced in the medical field to systematically close the gap between what we know and what we do (often referred to as the know-do gap) by identifying and addressing the barriers that slow or halt the uptake of proven health interventions and evidence-based practices.
Only recently is the importance of Implementation Science in the water sector being recognized. We are only making a start and have much ground to cover, but at least we have other sectors to learn from.
PS: If I seem overly critical of doctors, I am not. I am very grateful to doctors for their quick thinking in a very tiring and under-resourced situation. Many of the ER room patients were doctors themselves. My reflections above are only so we can hope doctors’ lives can be made easier, to strengthen their hand a bit more.