Outsiders Watch, Hospitals Reveal: A Three-Step Method That Turns Clinical Frustration Into Innovation

Walk through any European hospital today and you will encounter the same quiet crisis repeated in ward after ward: fewer staff caring for more patients, many of them elderly and living with multiple chronic conditions. Demographic change is not a distant forecast but a daily operational reality, and the burden lands squarely on the shoulders of physicians, nurses, administrators and logistics teams. A research team at the Technical University of Munich believes that the solutions to this crisis may already be hiding in plain sight, embedded in the everyday frustrations of the people who keep hospitals running. Their answer is not a new device or a piece of software, but a structured method for finding out exactly what needs fixing first.

In a study published in the International Journal of Computer Assisted Radiology and Surgery, Emily Spicker, Sonja Stabenow, Sidra Rashid, Lukas Bernhard, Charlotte Haid, Maximilian Berlet, Johannes Fottner and Dirk Wilhelm present a three-phase methodology designed to convert diffuse clinical complaints into a ranked, grouped and actionable list of innovation targets. The framework, which the authors describe through the three verbs observing, uncovering and pruning, was tested over three months at a German university hospital and produced 99 validated needs, of which 22 scored as particularly relevant, ultimately clustered into ten overarching themes. The promise is significant: a repeatable recipe that lets engineers and hospital managers see the hospital through the eyes of those who work in it, without drowning in anecdote.

The first phase, observing, deliberately breaks with convention. Rather than sending clinicians or seasoned process analysts into the wards, the team recruited four external observers with technical or economic backgrounds but no prior knowledge of hospital operations. The rationale draws on well-documented cognitive effects: experts habituate to inefficiency, while outsiders question routines that insiders have long stopped noticing. The observers covered four main areas of the hospital, including the nursing ward, the operating room tract, planning procedures and logistical operations, with supporting services such as administration and cleaning viewed in connection with these. Each observer recorded every activity deemed relevant to the daily workflow, along with the specific times or events that triggered it, and staff were asked to explain unclear processes, though only when no task involving a patient was underway.

Crucially, the observers did not simply accumulate notes. They modeled the event flow of the processes they watched and then verified those models with staff, treating any major discrepancy between the perceived and observed workflow as a signal of high process variation rather than as an error to be corrected. Weekly meetings among the observers and the study organizers enforced consistent note-taking, ensured sufficient overlap between the observed areas, and guarded against individual bias. The authors also emphasize the practical hygiene of the approach: observers dressed like employees, followed hospital hygiene regulations, and were introduced to all personnel in advance to secure consent and avoid confusion on the floor.

The second phase, uncovering, adapts a well-known tool from the world of medical device innovation. The Biodesign process proposed by Yock and colleagues at Stanford includes a needs-finding step that insists critical needs be identified before any solution is designed. The Munich team borrowed this logic and sharpened it for the hospital context, instructing observers to focus purely on when in a process a problem occurred, deliberately ignoring how or whether the problem might be avoided. This discipline is intended to reduce solution bias, the tendency to jump prematurely to fixes. Each identified issue was decomposed into three components: the problem itself, the population affected, and the intended outcome. The problem and outcome were then written into a structured statement of the form, a way to address the problem so that the outcome is achieved, with the affected population noted separately to maximize comparability across the growing list.

Staff validation formed the second step of the uncovering phase, confirming that the essence of each observed problem had been captured correctly and often generating further insight from the professional groups involved. By the end of this stage, the four observers had collected 130 raw needs, which the authors manually merged and filtered down to 99 distinct entries. That number alone illustrates the scale of latent demand for improvement in a modern hospital, but raw volume is useless without prioritization, and this is where the methodology’s most distinctive phase begins.

The pruning phase combines interdisciplinary workshops, staff surveys and mind mapping into a two-stage scoring system. In the first stage, six professionals from logistics, medicine, computer science and robotics gathered for four hours around six large sheets covering categories such as infrastructure, safe work environment, data handling, planning and communication. Each need was printed on a small slip of paper and rated on two criteria on a scale from one to five: the level of innovation required, weighted according to whether the project seeks rapidly available solutions or unsolved research problems, and feasibility given available resources such as team size and budget. The authors make the deliberate choice to double-weight feasibility in their scoring equation, reflecting the reality that hospital budgets are tight, regulations constrain experimentation, and overburdened staff will lose faith in any project that cannot realistically be delivered. Needs scoring eight or above survived this first cut, leaving 40 candidates for staff surveys.

In the second stage, the remaining needs were rated by healthcare professionals themselves in department-specific surveys, each need scored for frequency of occurrence, disruptiveness measured in added workflow time, and stress level. The team assigned equal weight to all three aspects for comparability, although they acknowledge that future refinements may weight them according to what staff actually value. A cutoff score of 15 reduced the list to 22 needs. Notably, the logistics department declined to participate because it was undergoing major restructuring, so the team devised an emergency workaround: logistical issues affecting nursing and operating wards were folded into those departments’ surveys, and the logistics and planning observers rated the remaining issues themselves, a solution the authors transparently flag as introducing higher bias. Finally, the surviving needs were grouped through iterative mind mapping, first by underlying cause and then by potential solution, producing ten thematic clusters ranging from documentation and communication to information access, material preparation and physical workload.

The results speak to more than the list itself. The authors report that the repeated involvement of staff across all three phases fostered a mutual understanding between clinicians and external researchers, boosting motivation and opening discussions about workflows that staff had never previously articulated. The presence of outside observers challenged patterns the staff themselves were no longer aware of, and in at least some cases prompted small, unprompted changes to routines before the formal study concluded. This aligns with a broader argument the authors make in their framing: employee-driven innovation in healthcare remains underrepresented in research and routinely lacks the organizational and technical resources to survive, while established approaches such as Lean Management have struggled against healthcare culture and the inherently complex, dynamic nature of clinical work. Process modeling languages like BPMN, the authors note, help formalize known processes but rarely surface the hidden systemic challenges that matter most.

The methodology is not without acknowledged limitations. It depends heavily on staff acceptance and participation, so the absence of any professional group creates knowledge gaps and skews final ratings, as the logistics case demonstrated. The process remains subjective, shaped by the personal biases of observers and participants, and not all stakeholders were consulted in the pruning stage, where interaction focused mainly on medical staff. Still, the authors argue that a multi-factor analysis of this kind is the only way to connect needs with actionable solutions, and they chart two directions for future work: deeper investigation of the underlying causes behind each identified need, and application of the method in other hospitals to compare needs across institutions and verify its effectiveness in different environments. If the approach travels well, it could offer hospitals a systematic bridge between the people who experience clinical friction daily and the engineers and managers equipped to eliminate it, turning the accumulated frustration of the wards into a ranked research agenda rather than a resigned shrug.

Subject of Research: A structured needs-finding methodology for analyzing hospital workflows and prioritizing healthcare innovation opportunities

Article Title: From clinical challenges to innovation: an idea finding approach for advancing healthcare

Article References: Spicker, E., Stabenow, S., Rashid, S., Bernhard, L., Haid, C., Berlet, M., Fottner, J., & Wilhelm, D. (2026). From clinical challenges to innovation: an idea finding approach for advancing healthcare. International Journal of Computer Assisted Radiology and Surgery. https://doi.org/10.1007/s11548-026-03789-8

Image Credits: AI Generated

DOI: 10.1007/s11548-026-03789-8

Keywords: needs finding, hospital workflow, healthcare innovation, process analysis, observational study, biodesign, workload reduction, hospital management, staff surveys, interdisciplinary collaboration, clinical documentation, medical technology