It was an unfortunate mishap. Things went badly wrong. Here’s how an Incident Command System can prevent nasty events happening again


Joshua Sharfstein, the former health commissioner of Baltimore, discusses the day a simple power failure led to embarrassment and dysfunction, and what they did to prevent it happening again.

It was a sweltering day during a hot Baltimore summer. Under the strain of everyone’s air conditioning demands, the grid faltered. One of the Health Department’s buildings lost power, and the 50 or so staff who worked there were trying to find out whether they should go home.

Some of the affected workers decided to call me, the health commissioner, on my cell phone. Others called the office, and a few even contacted City Hall directly. As my chief of staff and I tried to figure out what to do, we realised we had no plan for how to decide about closing one of our buildings, no established mechanism for distributing emergency news, and no way to update our workforce about plans for the next day. Then someone ran in and shouted that refrigerated vaccines in our overheated building needed to be moved immediately.

It took a couple of chaotic hours to get on top of the situation. My chief of staff and I all cancelled meetings, scrambled to track down key facts, and ironed out a plan. Eventually, we sent the staff home, set up an information telephone line for updates, and moved the vaccines. But we were embarrassed. A simple power failure had led to dysfunction.

We later met to figure out what had gone wrong.

First, no one had been in charge. The staff were hearing the full spectrum of possible instructions from different people, ranging from “whatever you do, don’t go home” to “leave now.” The ensuing confusion led to even more phone calls seeking clarification. Even as health commissioner, I did not have ready access to all the necessary information.

Second, the resources to address the problem were scattered across our organisation. The human resources office knew who worked in which buildings, the information technology office had all the email addresses, and facilities staff were needed to move the vaccines. None of the managers of these offices had been available when we needed them.

Third, in our confusion, we nearly missed a significant threat to the public health — the loss of our vaccine supply.

We decided that the next time would be different. We made a plan to switch management modes in the case of another power outage — from “normal,” governed by the usual organisation chart, to “emergency,” which would run under a different organisational arrangement called “incident command.”

As part of this structure, we would designate a single person to be in charge and act as the “incident commander.” Our top choice was the chief of staff, but if she were not available, a deputy commissioner would step in. We would communicate the switch to “emergency mode” by sending an email to the entire Health Department — an email we drafted in advance — or by calling staff through a phone tree.

The incident commander would make all the key decisions, consulting with me and City Hall as needed. She would also be able to deputise others to perform needed tasks, including drafting messages, setting up an information line, moving essential supplies like vaccines, and assessing for other public health risks. Everyone else would be able to keep doing their jobs as usual.

This new approach to power outages worked well — so well, in fact, that I cannot remember any of the other times our buildings lost power (and I’m told it happened from time to time).

A building with lost electrical power is trivial on the scale of public health crises. But the insight from this experience changed my approach to crisis management. I realised that crises often cannot be effectively managed through business as usual. As Boin and Hart have written:

“Unless organisations have been trained to recognise when their available repertoire does not suffice to deal with the crisis at hand, their operational routines will become more salient during crises (which creates a potential for further escalation of the crisis).”

Indeed, the ability to switch management gears is one reason why early recognition of a crisis is so important.

A useful management approach for responding to crises is the incident command system (ICS). Developed in the 1970s to coordinate efforts at the scenes of fires and other disasters, incident command is now the standard management structure recommended for a broad range of disasters by the Federal Emergency Management Agency (FEMA). There are dozens of books and courses that explain incident command in considerable detail, and FEMA offers free training courses online.

Once an agency has developed the ability to activate an incident command or a modified version of incident command, it is worth using it regularly — including to better manage everyday public health challenges. Doing so builds the muscles of an organisation in such areas as mobilising resources, communications, and decision-making under pressure.

Regular use of incident command can mean the difference in a crisis between a staff filled with excitement and confidence and a staff riddled with doubt or even dread.

Overview of the Incident Command System

Incident command differs fundamentally from the usual course of day-to-day management, by following four key principles.

Principle 1: one person should be present, in charge, and empowered to make decisions at all times

On a usual day, the agency director may be travelling and the deputy may be in meetings. It is common for there to be multiple pauses as decisions travel up and down the organisational chart. In a crisis, however, long waits for action can undermine the effectiveness of the response. Moreover, a power vacuum can lead to frustration and internal conflict at the very moment everyone needs to work together.

By contrast, under an incident command approach, “the command function must be clearly established from the beginning of the incident.” With someone in charge, and always available to make key decisions, the bulk of the response effort can focus on execution. The concept of a responsible person extends across the incident command management structure. According to FEMA, everyone involved in an incident command response “has a designated supervisor to whom he or she reports,” and the “manager at all levels must be able to control the actions of all personnel under their supervision.”

The incident commander, of course, does not have absolute power. A common question about incident command centres on the role of agency or elected leadership. What happens when the health commissioner or mayor, county executive or governor expects to participate in critical decisions? How can his or her role be squared with the job of incident commander?

An incident command structure avoids this problem by establishing a routine in which the incident commander arranges for comprehensive briefings for top officials at key decision points. The leadership can then participate in making important and timely decisions, which the incident commander can then implement.

Principle 2: the management structure should be designed based on the functions needed for the crisis

On a usual day, many different boxes on an agency’s organisational chart may touch a given project, including some whose contributions may be minimal. For example, an effort may involve program, fiscal, facilities, and IT staff — as well as communications, legislative affairs, and management. While there may be standard processes and procedures for bringing all of these roles together; these often take time. To respond rapidly during a crisis, it is often helpful to create a special emergency management structure that can prioritise key tasks.

Under incident command, the response structure can be scaled to the challenge. A small problem may require engaging only one person in the whole organisation — the incident commander. Larger crises require more elaborate structures, adding other roles as needed in two categories: command staff and general staff.

Command staff report directly to the incident commander. These positions include the press officer, who manages communications and coordinates with other communication officials in a Joint Information Center; the safety officer, who has the responsibility of minimising risks to the responders themselves; and the liaison officer, who handles coordination with other resources and levels of government.

General staff include those working in four sections:

  • Operations, responsible for “managing all tactical operations at an incident.” For example, a crisis response in public health may require the establishment of a vaccine clinic or a mobile unit to test for environmental contamination. An incident commander can create an operation section to oversee such activities.
  • Planning, which “collects situation and resource status information, evaluates it, and processes the information for use in developing action plans.” Crises that last for days and have many moving parts benefit greatly from a planning section devoted to putting together status reports and developing and revising an incident plan that provides coherence and direction to the effort.
  • Logistics, which provides for such needs as “facilities, transportation, communications, supplies, equipment maintenance and fuelling, food services (for responders), medical services (for responders).” An incident commander asks for a logistics section when there are major organisational requirements for a crisis response, such as feeding assembled staff or finding generators to deploy quickly.
  • Finance/Administrative, which “is responsible for managing all financial aspects of an incident.” The administrative section comes in handy later when the bill for a crisis is sent to the federal government for reimbursement.

Each of these key functions is led by a section leader, who can assemble a staff with specific roles and responsibilities, again depending on the scale of the crisis. In addition to these common units, incident command also allows commanders to establish specific goal-oriented teams that may be unique to the incident at hand.

Principle 3: the role is separate from the person

On a usual day, an agency role may be limited to a specific person, regardless of that person’s calendar. For example, if only William approves invoices, then invoices will sit in his inbox until William is at his desk.

Not so under incident command, which is designed for specific tasks to be completed when needed. So if the incident commander assigns the job of approving invoices to the leader of the finance/administrative team, then whoever is serving in that role will be ready to help. That may turn out to be William during one shift, Alice during the next shift, and Thomas overnight.

This concept holds across the incident command structure. For example, in a response that involves operations, the head of the operations section may change from time to time, but someone is always the head of the operations section. Indeed, an incident commander with a question for operations can simply say, “Find me the head of the operations section.” Some agencies ask their staff to wear fluorescent vests with their incident command titles in huge letters, designed to make identification of their roles that much easier. Passing on the vests at the end of a shift is accompanied by “a briefing that captures all essential functions for continuing safe and effective operations.”

Some people in every organization truly hate the concept of separating the role from the person. After all, if I’m the expert in epidemiology, why would I let someone else take over data analysis, even for a moment? But the truth is that organizations cannot be dependent on any one person to function in crisis. Providing training in advance for each role is a critical and essential task of crisis preparation. Once the response has started, empowered staff need to rise to the occasion.

Principle 4: management by objective

In the usual course of business, daily agency action is generally guided by custom and habit, based on years of evolution of specific programs and initiatives. In a crisis, however, circumstances can change rapidly, and it helps to develop and revise a written plan with clear objectives.

Under an incident command structure, this task usually falls to the planning section. With input from key experts inside and outside of the agency, the planning section drafts and revises key short- and medium-term objectives, provides status updates based on the latest data, and recommends changes to strategy.

After a crisis ends, the planning section can lead a session (called by some a “hot wash”) to review what went right and what went wrong in the response and develop an after-action report. Such a cycle of review can reinforce a culture of continuous improvement.

The utility of such a deliberate approach crisis management has been demonstrated time and again, from localised weather events in a single county to disasters affecting a nation’s future.

This is an extract of ‘Public Health Crisis Survival Guide: Leadership and Management in Trying Times’ by Joshua Sharfstein, former health commissioner of Baltimore and currently of Johns Hopkins Bloomberg School of Public Health. Published by Oxford University Press, June 2018.

About the author
Premium

The essential resource for effective public sector leaders

Check out the Latest