Reinvigorating Proportionality to Protect Hospitals During War

Over the past five years, the number of attacks on health facilities and personnel in armed conflict has tripled, with the most incidents in Myanmar, Gaza, Sudan, and Ukraine. The attacks inflict massive harm on patients, medical staff, hospital functioning, and the local population, including death, suffering, exacerbation of disease, and destruction of key facility functions and capacity to treat patients. 

These incidents have occurred despite the special protections that exist for hospitals and other medical facilities under international humanitarian law. Perpetrators are increasingly suggesting that these targeted facilities have lost their protected status by committing, outside their humanitarian functions, “acts harmful to the enemy.” Those acts can include employing the facility to fire or store weapons, house fighters for reasons other than medical care, or operate an intelligence unit or command center.

Once protection is lost, hospitals can be attacked—but they are still subject to a number of safeguards. These include a determination that the hospital has become a legitimate military object, that a warning to cease military use has been issued but not heeded, and that general obligations of precautions to avoid or at least minimize harm to civilians and proportionality in attacks have been met.

But the use of proportionality as a means of avoiding or limiting harm in cases of alleged military use of a hospital is rarely discussed. A reinvigorated analysis by military forces can reaffirm proportionality’s application to hospitals during times of war and prevent brutal attacks that result in unnecessary and sometimes fatal harms to patients in the facility and people who need hospital-level care.

Failures of Protection

In the war beginning in 2023 in the Gaza Strip, as of October 2025 there were more than 400 incidents of violence damaging health facilities, including a third from Israel Defense Forces (IDF) airstrikes. During the war, 34 of Gaza’s 36 hospitals were damaged. The IDF frequently claimed that its attacks were justified based on evidence of Hamas using hospitals for military operations—a justification sometimes backed by the Pentagon. There is evidence that, in the 2014 conflict in Gaza, Hamas conducted interrogations in Al-Shifa Hospital, and in the current one its tunnel network extended under or near some hospitals. Many of Israel’s assertions about Hamas operations in hospitals, however, have been contested by hospital administrators and staff, international medical volunteers, and media and UN human rights investigators.

It is usually impossible to resolve these factual disputes in the midst of war. But even if the allegations are accepted as true, patients and medical personnel in the facilities retain their immunity from attack, and attackers must follow specific legal requirements to protect them and the essential functions of the hospital. One requirement, unique to health facilities, is that before initiating an assault, the attacker must provide a warning demanding the cessation of harmful acts and, in appropriate cases, specify a reasonable time to respond. The attack may commence only if the warning goes unheeded. 

However, the warning requirement has been largely ineffective in preventing attacks on hospitals. The required warnings are rarely given, and if they are, attacks are not delayed while awaiting a response—or they are not responded to entirely. The precautions requirement, if taken seriously, is more widely recognized as having potential to avoid harm to patients, staff, and hospital functioning. The International Committee of the Red Cross (ICRC), in conjunction with 29 state militaries, has offered practical measures to meet this requirement in connection with health care facilities.

These can include developing communication and coordination mechanisms with medical administrators; limiting authorization of operations against hospitals to those at a high level of command; avoiding use of highly explosive weapons including missiles, bombs, and artillery against hospitals; taking concrete steps to ensure continuity of electricity, oxygen, and water for life support during assault; preserving of core hospital functions such as intensive care units; and planning for safe evacuation of the most vulnerable patients. Additionally, tactical patience can avoid significant harm.

But the extent to which such steps are taken is contested. In Gaza, the IDF has claimed to have taken some of these measures in hospital attacks, particularly evacuation of patients, though others, including me, have found them absent or carried out under horrifying circumstances of lack of electricity and water, and unsafe routes. The precautions rule also grants enormous discretion to commanders and troops when weighing feasibility, and does not address whether an attack should be avoided in the first place or suspended during its execution because of likelihood of excessive harm to civilians.

Militaries must take obligations of warnings and precautions more seriously but also elevate the rigor of proportionality assessments, as doing so can mean foregoing attacks against hospitals altogether or severely limit their severity.  

Proportionality: A Means to Avoid Violence Against Hospitals

The rule of proportionality in attacks prohibits “launch[ing] an attack which may be expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects, or a combination thereof, which would be excessive in relation to the concrete and direct military advantage anticipated.” It also requires that an attack be canceled or suspended if it becomes apparent before or during an assault that the anticipated civilian harm outweighs the anticipated military advantage.

This obligation could be the most protective of all, and most straightforward to implement in military operations against health facilities. Military forces don’t have to disclose their proportionality analysis, but in public messaging in defense of attacks on hospitals, it has been largely ignored, which suggests little attention has been paid to it.

Though an intentional violation of the proportionality obligation is a war crime under the Rome Statute, the rule is something of a stepchild of the law. One reason is the difficulty of balancing two unlike things—excessive civilian harm and concrete military advantage. Further, proportionality is not judged by the results of the attack, but by the commander’s assessment, before or during an attack—based on information reasonably available at the time—of the expected level of civilian harm caused in the attack in relation to the anticipated military objective.

It is unsurprising, then, that there have been only a handful of cases in international tribunals regarding proportionality. When military forces claim that a hospital attack meets the proportionality standard, they rarely explain the basis for that determination. Further, as I have previously argued, some militaries, including Israel, have erroneously diluted the force of the proportionality rule by suggesting that anticipated military advantage includes any step toward the overall goal of winning the war, rather than a “concrete and direct” advantage to be achieved in the specific military operation as a whole, as the rule requires. This leads to the claim that incidental harm to civilians and civilian objects will almost never outweigh the military advantage gained.

Despite these challenges, the proportionality rule has the potential to be a powerful mechanism for preventing or mitigating the grievous harms to patients, staff, and facility functioning in assaults on hospitals in armed conflict, including instances where such attacks are justified by allegations of their use to commit acts harmful to the enemy. As Michael Schmitt puts it succinctly, the proportionality rule means that in some circumstances “militarily necessary gains may not be pursued because the civilian cost is too high.” He notes, too, that if there is to be an attack, proportionality controls “the nature and extent of the forcible response.”

Assessing Proportionality in Hospital Assaults

Applications of proportionality are based on a military commander’s assessment of the foreseeable injuries or deaths to civilians or damage to civilian objects during the attack, and the military advantage anticipated.

In determining excessive harm to civilians and civilian objects in hospital attacks, unique considerations apply. First, the special protections of the wounded and sick and health personnel, as well as the essential functions of a hospital, can tilt the balance toward foregoing the attack or limiting its scope. Second, as the U.S. Law of War Manual explains, expected damage to civilian objects must be given “greater consideration” when there is a risk of harm to civilians inside them (Section 5.12.1.1). Third, the manual notes that if feasible precautions to avoid or minimize harm have not been taken, an inference of excessiveness is stronger (Section 5.12.3.3).

Depending on how the attack is carried out and the weapons employed, expected deaths and injuries—including exacerbation of medical conditions—may be a product of firing on the facility, cutting off power during the assault that is essential for life support for vulnerable patients, impeding transport of supplies and medication to the facility, depriving patients of water and food, or forcing hurried and dangerous evacuation. The facility and staff may lose the ability to treat injured patients and bystanders in or outside the facility. The attack may also be expected to damage key functions of the facility, including intensive care, surgery, and safe childbirth. As a result of these conditions, moreover, some patients are likely to be released before it is medically appropriate to do so. The facility may also be expected to suffer significantly reduced capacity to meet the needs of both current patients and the local population.

In Gaza, there is extensive evidence of these excessive—and forseeable—harms. Patients have been killed by explosive weapons fired into or at hospitals, died as life support ended, and suffered exacerbations of illness. Medical staff have also been killed in assaults. Essential functions of hospitals have been severely damaged or destroyed even as the need for these capacities has increased dramatically. Some forms of treatment, including for cancer, traumatic injuries, kidney disease, and infectious disease, have halted as a result of attacks.

There are different views on whether longer term, and sometimes indirect, effects of attacks must be considered in the proportionality analysis. The ICRC and some states have interpreted proportionality as requiring consideration of these indirect, reverberating harms, while others, reflected in the U.S. Law of Manual, require consideration only of “immediate and direct harms” (Section 5.12.1.3). I side with those who argue that indirect effects must be considered, but in the case of attacks affecting hospitals, even the narrower view recognizes that direct harms can occur along a chain of likely causality. For example, in the Law of War Manual, the same section that mandates consideration only of direct effects explains that an attack on a power plant that could be expected to result in loss of power to a connected hospital is a direct and immediate harm. By the same reasoning, an attack on a hospital that results in the population being denied access to essential health services must be taken into account. The example also shows that general knowledge of likely harms, based on past experience or empirical data, is necessarily part of the assessment of expected civilian harm. 

These expected, excessive harms to access and services often increase during the course of war, further shifting the balance in the proportionality calculus. As Schmitt points out, the medical crisis in Gaza has deepened over time, with the harms to the civilian population becoming increasingly excessive compared to the military advantage gained. As a result, the proportionality calculus “may yield less military advantage and risk greater collateral impact” as a conflict progresses, requiring constant reevaluation.

The military advantage side of the proportionality assessment also has implications for what parts of a hospital may be assaulted, and by what means. As the ICRC commentaries to the Geneva Conventions explain (paragraph 1860), because of the special protection of the wounded and sick, only the part or parts of the hospital used for acts harmful to an enemy are military objectives and may be targeted. The U.S. Law of War Manual notes, for example, that a proportionate response to an enemy rifleman firing from a hospital window would warrant a response against the rifleman only, rather than the destruction of the entire hospital (Section 7.17.1.2).

Similarly, in 2005 Israel’s High Court of Justice considered the application of proportionality in the context of targeted killing of alleged terrorists. In its exposition of the application of the proportionality requirement, it provided a hypothetical example of a sniper shooting at civilians or soldiers from his porch. The opinion noted a clear military advantage in killing him, but suggested that the proportionality calculus would yield different results depending on the type of military response employed. It explained that the death of a civilian neighbor in shooting at the sniper would not be excessive in relation to the advantage in killing the sniper. By contrast, it would not be proportional “if the building were bombed from the air and scores of its residents and passersby were harmed.”

The same reasoning should prohibit a bombing, missile, or shelling attack on a hospital on account of its having a closet for storing weapons, or a room used as a command center. Indeed, only in the most rare or extreme circumstances could the military advantage gained from employing such weapons outweigh the excessive harm to patients, staff, and hospital functioning caused by them. Similarly, a ground assault on a hospital in the event of prohibited military uses of a part of it should be severely circumscribed, else the harm to civilians would exceed the military advantage gained.

In Gaza, the IDF has mostly alleged that Hamas stored weapons, placed tunnels underneath hospitals, established entry points from a hospital to a command center, and/or employed hospitals as “hideouts” for unspecified numbers of Hamas fighters. In one case, it cited (incorrectly, as it turned out) Hamas employing a camera on a balcony of a hospital Hamas as an observation post as justification for multiple tank shellings of the hospital, killing 22 people. It is dubious whether the IDF conducted any proportionality assessment or that a proper one would assess that the military advantage outweighed the excessive expected harm to civilians in that case. In many other instances, too, the IDF employed explosive weapons, including missiles, artillery, and bombs, against hospitals—instances that cannot withstand proper proportionality scrutiny.

Gillard has offered very specific recommendations on how to carry out proportionality assessments generally. These could go a long way in avoiding or mitigating harms to hospitals and their patients and staff. To begin, detailed requirements for meeting the proportionality rule should be included in military doctrine, manuals, and rules of engagement, both in development and execution of planned attacks and in dynamic attacks. Militaries should establish requirements for information gathering and analysis of civilian harm in hospital attacks from publicly available information, intelligence sources, humanitarian and medical providers, and knowledge of the consequences of prior attacks. They should specifically address when attacks should be suspended, and train frontline troops and commanders in making real-time judgments. A thorough assessment of the military advantage anticipated is also essential. Finally, formal reviews  of proportionality assessments and  lessons learned from them can enhance compliance with proportionality imperatives.

In addition to Gillard’s proposals, militaries should adopt detailed rules for proportionality assessments in the case of hospital attacks. · 

*   *   *

Taking proportionality seriously can save lives in hospital attacks. In the battle of Fallujah in 2004, American and Iraqi forces assessed that the main hospital, which was strategically located, was being used by armed insurgents for military activities. The forces carefully planned a ground assault that involved breaking into the hospital and capturing it. They performed the operation without any deaths or injuries (other than an accident in which an Iraqi soldier shot himself in the leg), and maintained the ability of the hospital to function normally. The ICRC’s guidance on protecting hospitals in military operations, noted above, includes recommendations on how to conduct hospital assaults that can achieve military objectives without unduly interfering with the hospital’s functioning and avoiding harm to patients and staff.

One of the principal limitations of the proportionality rule—that it is based on expectations of harm, not the actual results of military actions—turns out to be a major strength in avoiding or mitigating the horrors befalling wounded and sick people in assaults on hospitals in war. The rule’s requirements must be taken seriously, necessitating a rigorous assessment of expected excessive harm and a serious evaluation of the often-limited military advantage sought. There is almost always abundant information available to perform such as assessment before—and even during—military operations, enabling a proper judgment on whether and how to act, and saving countless lives.