The ol’ suggestion box is a fixture of a number of offices. Does anything ever get put in there, I wondered? Are any of the suggestions useful? Do any of them get implemented? Well, thanks to a recent post in the HBS Working Knowledge blog, I don’t have to wonder any longer:
Bumping up against accepted theories in process improvement, a new research paper from Harvard Business School questions the value of prioritizing problems identified by frontline employees.
Citing a hospital safety improvement program based on employee suggestions, researchers Anita L. Tucker and Sara J. Singer show that the commonly accepted “analysis” approach, wherein great attention is given to identifying and prioritizing a large number of problems, is not associated with success.
Instead, an “action” approach is preferable. According to their research, hospital units that focused on fixing easy-to-solve problems had greater improvement in safety climate than those focused on identifying a bunch of hard-to-solve ones.
“To our knowledge, this tradeoff between analysis and action in process improvement programs has not been empirically examined,” write Tucker and Singer in their June 2012 working paper, Key Drivers of Successful Implementation of an Employee Suggestion-Driven Improvement Program. Tucker is an associate professor in the Technology and Operations Management unit, and the Marvin Bower Fellow at HBS. Singer is an assistant professor of health care management and policy at the Harvard School of Public Health.
Trained as an industrial engineer, Tucker is interested in the perspective of frontline workers in productivity and process improvement, including how internal supply chains and other processes get them what they need to do their jobs. Her interest in health care has led her to study nurses, who, as direct care providers, are at the center of the web of supply chains of equipment, supplies, medications, and even physicians.
Singer received a grant from the Agency for Healthcare Research and Quality to develop an instrument to measure safety climate in hospitals. One condition of funding was that she also create a tool to improve the safety climate and encourage senior managers to become more engaged in hospital safety initiatives.
Fixing by walking around
The program the researchers tested was modeled on Allan Frankel’s “Leadership WalkRounds,” which has been shown to improve safety in various medical facilities. His hypothesis was that “management by walking around” (MBWA) enabled senior leaders to put together a task list of issues that, if addressed appropriately, would make for safer conditions.
Tucker and Singer randomly selected 20 hospitals to conduct an MBWA-based program for 18 months, starting in 2005. A total of 58 departments participated, or roughly three departments from each hospital. Most frequently involved were the emergency department, medical/surgical, operating room, and post-anesthesia care units, but some hospitals also included labs, the pharmacy, or medical records.
The senior leadership teams engaged in MWRA were usually composed of chief executive, operating, nursing, and medical officers. Each officer went out for roughly an hour at a time and was encouraged to obtain feedback from nurses, doctors, respiratory therapists, and administrators. Later, the officers held large staff meetings in each unit to gather more information and discuss safety concerns, which previous research suggests increases such programs’ positive impact on culture. The process generated a tick list of approximately 20 items, which was evaluated by the senior and unit leadership together in a debriefing.
Is it a priority?
The commonly accepted wisdom in the process-improvement literature—including Frankel’s work—is that since not all problems identified by frontline staff are of equal importance, it is worth time and money to perform extensive prioritization. Suggestions that will produce the greatest improvement should be focused on first. Then, as Tucker says, “you work down the list until you run out of either time or money.” But Tucker and Singer’s findings diverge from previous research and theory.
About half the hospitals that participated in their study chose to do some kind of prioritization—the analysis approach. Using a system similar to that adopted in Frankel’s work, Tucker and Singer assigned each suggestion a priority score based on importance, likelihood, severity, and difficulty of correction.
Identifying the hospitals with the most improved safety scores turned up some surprising findings. For one, management intervention helps. “Regardless of how we analyzed the data (at the level of the respondent, the unit, or the hospital), where a higher percentage of the problems were assigned to a senior leader for action, there was a greater improvement in safety climate.”
The quality-improvement literature devotes a great deal of time to the idea that because frontline staff is empowered to fix problems, senior leaders should sometimes cede responsibility. Thus, Tucker and Singer’s paper puzzled some readers, who found it counterintuitive that improvements came when senior leaders seemed to be taking away employee power.
Tucker has a different perspective. “I believe it was not that senior leaders were solving problems, but that they were just making sure things didn’t fall through the cracks,” she says.
Tucker also explains this finding in terms of “boundary spanning.” Nurses are at the far end of an internal supply chain. Even if they discover a gap between what the supply chain is providing and what the patient needs, they usually don’t have the authority or knowledge to go back to those supply departments and fix the problem; a higher-level person needs to be involved.
“This finding tells us that process improvement in hospitals will require people to work across departmental boundaries, where the problems happen, rather than within a particular department,” Tucker says.
Analysis versus action
The next thing Tucker and Singer looked at was analysis versus action in problem solving, which describes how an organization spends its resources. The analysis approach holds that by identifying lots of issues, a high-value problem will turn up, which if solved will yield a disproportionately high benefit. However, the researchers were in for another revelation.
“We found that the analysis approach was not associated with success at all, which surprised us, since it’s so ingrained in the process improvement literature,” says Tucker. “We found instead that the action approach was more successful: fix what you know about first. Units that solved a higher percentage of easy-to-solve problems—’low-hanging fruit’—showed greater improvement in safety climate.”
One reason for this is that employees are more likely to buy into improvement programs when their suggestions are actually implemented. The analysis approach yields a large number of potential fixes, but only a frustratingly few, from the employee’s perspective, are acted upon. And the analysis process can be time-consuming.
The paper cites a hospital where lab results took a long time to be completed, causing patient backups. Managers held a safety forum to surface possible solutions. “We observed the manager spent the entire time getting staff input on prioritizing the items—such as severity, frequency of occurrence, and ease of solution—leaving no time to discuss how the issues might be resolved,” the authors write.
Conversely, another hospital identified that a medication room was too small for more than one nurse to work in at a time, delaying patient care. Senior managers discussed the issue with staff and they collectively made a plan to move the medication room to a larger space.
In addition to the importance of senior management cooperation and more doing than analyzing, the findings yield several other practical applications. For one, process improvement appears to be like a muscle, Tucker says. The more you exercise it the stronger it becomes.
“Especially with boundary-crossing changes, practicing and spreading that skill set through the organization can solve more and bigger problems,” Tucker says. “They’re not going to get better by picking the right problem. They’re going to get better by becoming better problem-solvers.”
The findings mesh well with Tucker’s recent readings on decision-making biases.
“Basically, people place more value on a good outcome today than a better outcome for which they would have to wait for two weeks,” she says. “That’s why I think process improvement needs to be reconceived as today’s work. It’s not something discretionary that will make my job easier next week or two months from now; I do it today because I have to do it today to get my job done.”
Tucker experienced this in other research where she studied nurses confronted with a problem. Would they speak up and try to solve the issue permanently over time—the better outcome—or create an expedient work around? The nurses overwhelmingly chose to work around a problem, because that is what allows them to get their very demanding jobs done in the most efficient way.
Tucker’s future research will continue focus on how human behavior intersects with process improvement. “I’ve come out of this with the goal of designing process improvement systems that take into account how people really behave, not how we want them to behave,” she says.