Case Study 1: Medicine and Foreign Policy -- Two Theaters of Iatrogenic Harm

"We have met the enemy and he is us." -- Walt Kelly, Pogo (1970)

Two Systems, One Pattern

This case study examines iatrogenesis in two domains that could hardly seem more different: medicine and foreign policy. One operates on individual bodies; the other operates on geopolitical systems. One is practiced by physicians trained in biochemistry and anatomy; the other is practiced by diplomats and generals trained in political science and strategy. One aims to heal the sick; the other aims to secure national interests.

Yet the structural pattern of harm is identical. In both domains, well-intentioned professionals intervene in complex systems they do not fully understand, using models that capture some variables and miss others, producing consequences that diverge -- often catastrophically -- from their intentions. The parallel is not metaphorical. It is structural, and understanding the structure reveals why iatrogenesis is so persistent and so difficult to prevent.

Part I: The Opioid Crisis as Medical Iatrogenesis

The Problem That Demanded a Solution

In the early 1990s, a genuine problem existed in American medicine: chronic pain was undertreated. Patients with severe, persistent pain -- from cancer, from injuries, from degenerative diseases -- were suffering unnecessarily because physicians were reluctant to prescribe opioid painkillers. This reluctance had historical roots: the medical profession had long been cautious about opioids because of their addictive potential. Pain management specialists argued, with justification, that this caution had become excessive. Patients were suffering. Something needed to be done.

The medical establishment responded. In 1996, the American Pain Society introduced the concept of "pain as the fifth vital sign" -- arguing that pain should be measured and treated as routinely as blood pressure, heart rate, temperature, and respiratory rate. The Joint Commission on Accreditation of Healthcare Organizations (JCAHO) adopted pain management standards that effectively required hospitals to assess and treat pain aggressively. Pharmaceutical companies, sensing a market opportunity, promoted new opioid formulations -- most notoriously OxyContin, manufactured by Purdue Pharma -- as safe and effective for chronic pain with minimal addiction risk.

The intervention was comprehensive: new clinical guidelines, new regulatory standards, new pharmaceutical products, new cultural expectations. Physicians who had been cautious about opioids were now pressured to prescribe them. Hospitals that had tolerated undertreated pain were now penalized for it. The entire system shifted toward more aggressive pain treatment.

The Iatrogenic Cascade

The intervention treated the original problem. Pain was managed more aggressively. Patients reported lower pain scores. The measurable metrics improved.

But the intervention also initiated a cascade of iatrogenic harm that would become the worst drug crisis in American history.

First-order effect (intended): More patients received opioid prescriptions for pain. Pain scores improved.

Second-order effect (unintended): A significant fraction of patients prescribed opioids developed dependence. The drugs were more addictive than the pharmaceutical companies had claimed and more addictive than the clinical guidelines had assumed. Patients who had been prescribed opioids for a legitimate injury found themselves unable to stop when the injury healed.

Third-order effect (unintended): As prescriptions were eventually tightened in response to the growing addiction crisis, patients with established opioid dependence lost access to pharmaceutical opioids and turned to illicit sources: heroin and, eventually, illicitly manufactured fentanyl. The supply shifted from regulated to unregulated, from known dosages to unknown dosages, from pharmaceutical purity to street-level adulteration.

Fourth-order effect (unintended): Fentanyl, approximately fifty times more potent than heroin, produced a wave of overdose deaths that dwarfed everything that had come before. The overdose death rate in the United States climbed from approximately 17,000 in 1999 to over 106,000 in 2021.

The full cost of the opioid crisis is difficult to calculate. Estimates of the total economic cost exceed one trillion dollars. Over 500,000 Americans have died of opioid overdoses since 1999. Millions more have experienced addiction, family disruption, job loss, and incarceration. Rural communities that were already economically distressed were devastated.

The Structural Analysis

The opioid crisis is a textbook case of iatrogenesis because every link in the causal chain connects an intervention to a harm:

1. The intervention to treat chronic pain more aggressively was a response to a real problem (undertreated pain).
2. The intervention was implemented through institutional mechanisms (clinical guidelines, accreditation standards, pharmaceutical marketing) that amplified its reach far beyond what the original problem required.
3. The intervention's benefits were measured (pain scores improved) while its costs were not measured (addiction rates were not tracked by the same institutions that tracked pain scores).
4. When the costs became visible, the corrective intervention (restricting prescriptions) created a new problem (patients switching to illicit drugs) that was worse than the problem it addressed.
5. Each corrective intervention generated its own iatrogenic effects, creating a cascade of interventions and harms.

The McNamara Fallacy was in full operation: pain was measured (and managed); addiction was not measured (and was ignored). The Goodhart's Law dynamic was active: pain scores became targets that incentivized prescribing regardless of addiction risk. The intervention bias was pervasive: at every stage, the pressure was to "do something" -- prescribe more, restrict more, enforce more -- rather than to pause, assess, and consider whether the intervention itself was the problem.

What Would Via Negativa Have Looked Like?

A via negativa approach to chronic pain management might have looked like this:

Instead of adding a new class of interventions (aggressive opioid prescribing), the medical system could have focused on removing obstacles to existing pain management strategies: better access to physical therapy, removal of insurance barriers to non-pharmacological pain management, reduction of the surgical interventions that often initiated the pain cycle in the first place. Rather than adding opioids, subtract the conditions that create chronic pain.

This approach would have been less dramatic, less measurable, and less profitable. It would not have produced a new blockbuster drug or a simple metric (pain as the fifth vital sign) that could be tracked on a hospital report card. It would have been harder to implement, harder to measure, and harder to build a career around.

It would also not have killed 500,000 people.

Part II: The War on Terror as Foreign Policy Iatrogenesis

The Problem That Demanded a Solution

On September 11, 2001, al-Qaeda terrorists hijacked four commercial aircraft and killed nearly 3,000 people in the deadliest terrorist attack in history. The United States faced a genuine threat: a terrorist organization with global reach, ideological motivation, and demonstrated capability. A response was necessary.

The response was massive. The United States invaded Afghanistan in October 2001 to topple the Taliban government that had sheltered al-Qaeda. It invaded Iraq in March 2003, claiming that Iraq possessed weapons of mass destruction and posed an imminent threat. It launched a global "War on Terror" that expanded military operations, intelligence gathering, and covert action across dozens of countries.

The initial military operations achieved their immediate objectives. The Taliban government fell within weeks. Al-Qaeda's leadership was scattered. Saddam Hussein was captured and executed. By conventional military measures, the interventions were successful.

The Iatrogenic Cascade

The interventions also initiated a cascade of consequences that, measured against the stated goals, constituted one of the most consequential instances of foreign policy iatrogenesis in modern history.

Afghanistan: The invasion toppled the Taliban but failed to establish a stable replacement government. American and allied forces remained in Afghanistan for twenty years, fighting an insurgency that the invasion itself had energized. When the United States withdrew in August 2021, the Taliban retook the country within days, returning to power with two decades of additional grievance, hardened combat experience, and a vast arsenal of American-made military equipment left behind. The twenty-year intervention cost an estimated $2.3 trillion and achieved, in strategic terms, the restoration of approximately the same situation that existed before the intervention.

Iraq: The invasion eliminated a dictator but created a power vacuum that detonated decades of suppressed sectarian conflict. The dissolution of the Iraqi army released hundreds of thousands of armed, trained men into a society with no institutions to absorb them. Many of these men formed the core of the insurgency and, eventually, of the Islamic State. At its peak in 2014, ISIS controlled territory across Iraq and Syria, operated as a proto-state with its own tax system, judicial system, and military, and inspired or directed terrorist attacks worldwide. The intervention designed to eliminate a threat had created a far more dangerous one.

Regional destabilization: The Iraq War destabilized the broader Middle East. The power vacuum in Iraq enhanced Iranian influence in the region -- the opposite of what the intervention's architects intended. The perception of American overreach strengthened anti-Western sentiment. The imagery from Abu Ghraib prison and civilian casualties from drone strikes became recruiting tools for terrorist organizations worldwide.

Domestic costs: The War on Terror consumed an estimated $8 trillion in direct spending and produced over 900,000 deaths across all theaters, including over 7,000 American military personnel and hundreds of thousands of Afghan, Iraqi, Pakistani, and other civilians. The domestic surveillance apparatus expanded massively, raising civil liberties concerns that persist to this day.

The Intervention Spiral in Full

The War on Terror illustrates the intervention spiral with particular clarity:

1. Initial problem: A terrorist attack kills 3,000 people.
2. Intervention: Military invasions of Afghanistan and Iraq.
3. Iatrogenic effect: Power vacuums, sectarian conflict, radicalization.
4. New problem: ISIS emerges, more dangerous than al-Qaeda.
5. New intervention: Military campaign against ISIS.
6. New iatrogenic effect: Further destabilization, civilian casualties, refugee crisis.
7. New problem: Regional instability, European refugee crisis, continued terrorism.
8. New intervention: Continued military presence, drone campaigns, counterterrorism operations.

Each round of intervention produced consequences that appeared to justify further intervention. Each round of further intervention produced new consequences. The spiral continued for two decades. The original problem (a terrorist organization with several hundred core members operating from caves in Afghanistan) metastasized into a generalized regional conflagration involving multiple countries, millions of displaced people, and terrorist organizations more numerous and more capable than the original threat.

The Structural Isomorphism

The Common Structure

Both cases share a structure that makes iatrogenesis not just possible but probable:

Genuine urgency. In both cases, a real problem existed that demanded attention. Undertreated pain was real. The 9/11 attacks were real. The pressure to act was legitimate. Iatrogenesis does not require that the original problem be imaginary. It requires only that the intervention be poorly calibrated to the system's complexity.

Incomplete models. In both cases, the interveners operated with models that captured some variables and missed others. The pain management model captured pain intensity and missed addiction vulnerability. The military planning model captured conventional military capability and missed sectarian dynamics, cultural factors, and the radicalization effects of occupation.

Measurable benefits, unmeasurable costs. In both cases, the intervention's benefits were tracked by the institutions that implemented them (pain scores, enemy combatants eliminated) while the costs were not (addiction rates, civilian radicalization). The McNamara Fallacy ensured that the intervention looked successful by the metrics being tracked, even as it was failing by the metrics that were not.

The spiral. In both cases, the iatrogenic consequences of the initial intervention were addressed not by reconsidering the intervention but by adding more intervention. Prescription restrictions led to more enforcement. Military setbacks led to more military force. Each additional intervention produced its own iatrogenic effects, sustaining the spiral.

Difficulty of reversal. In both cases, the intervention created constituencies that resisted reversal. The pharmaceutical industry profited from opioid sales. The military-industrial complex profited from the War on Terror. Pain management guidelines became institutionalized. Counterterrorism became a permanent bureaucracy. Path dependence locked in the iatrogenic intervention.

Questions for Reflection

1. The opioid crisis began with a genuine problem (undertreated pain) and a reasonable response (more aggressive pain management). At what point did the intervention become iatrogenic? Could the iatrogenic harm have been predicted before it occurred? What information would have been needed?

2. The War on Terror illustrates the intervention spiral: each round of intervention produces consequences that appear to justify further intervention. How would you break this spiral? What would a policy of via negativa look like in the context of counterterrorism?

3. Both cases exhibit the McNamara Fallacy: measuring what is easy to measure and ignoring what is not. Design a measurement system for either case that would have captured the iatrogenic costs as well as the benefits. What obstacles would you face in implementing this system?

4. In both cases, corrective interventions (restricting prescriptions, fighting ISIS) produced their own iatrogenic effects. Is there a general principle for designing corrective interventions that do not themselves cause iatrogenic harm? Or is iatrogenesis inevitable whenever complex systems are intervened upon?

5. Both cases created institutional constituencies that resisted reversal of the intervention. How does this connect to the path dependence concept (preview of Chapter 21)? Is there a way to design interventions that do not create constituencies committed to their continuation?