THE QUESTION
Why do medication errors keep happening?
The causes, deepest first. Disagreements stay on the map.
closed · back to Actions
Viewing as guest. Enter a handle to join and take part.
The map
Left to right: deepest causes first; outlined green boxes are the roots. Only direct links are drawn — a link implied by a path is on the record, not in the picture. Dashed red is contested. Hover for full text.
How this dialogue measures
why
The indicators the method's own practitioners use to judge a dialogue's quality (Laouris & Metcalf 2024, Table 7), computed from this record so it can be compared with the hundreds of face-to-face and virtual dialogues behind their baselines. Spreadthink is the share of ideas that got at least one vote: 100% means everyone voted only for their own, and it usually settles to 25–45% once meaning has been shared. Situational complexity falls as more ideas are structured.
ideas (N)
28 usual: 50–100
28 usual: 50–100
with 1+ vote / 2+ votes
24 / 22
24 / 22
spreadthink
82.6% usual: 25–55
82.6% usual: 25–55
demosensus
17.4%
17.4%
clusters (dimensions)
5
5
cards in a cluster
35.7%
35.7%
structured / levels / links
20 / 3 / 12
20 / 3 / 12
situational complexity
0.55
0.55
minds moved
—
—
cards hidden by the steward
0
0
pairs contested
0
0
The merged map
why
Every group built its own map from its own wall; these are folded into one. The same claim raised in two groups is one item, and each connection shows how many groups found it. Where groups found opposite things, that is published as a conflict.
ROOT CAUSES
◆ 88f10f0
◆ 626fd23
◆ 61aa254
◆ 35cf04a
◆ 108bae0
Connections several groups found (12 of 12)
| Infusion pumps on the ward come from three manufacturers | → | Agency staff do not know the ward's conventions | 1 group |
| Pharmacy reviews charts a day after the first dose was given | → | Night shift has one nurse for the whole bay | 1 group |
| Near misses are not reported because the form takes twenty minutes | → | The same look-alike drugs are stored next to each other | 1 group |
| Allergy information is in a different system from the prescribing one | → | The same look-alike drugs are stored next to each other | 1 group |
| Allergy information is in a different system from the prescribing one | → | Alerts in the system are so frequent that they are clicked through | 1 group |
| Double-checking is signed without a second person actually looking | → | Discharge medication lists disagree with the inpatient chart | 1 group |
| Patients are moved between wards mid-treatment and the chart lags behi | → | Discharge medication lists disagree with the inpatient chart | 1 group |
| Night shift has one nurse for the whole bay | → | Interruptions during the drug round are constant | 1 group |
| Errors are discussed as individual mistakes, not as patterns | → | Near misses are not reported because the form takes twenty minutes | 1 group |
| Errors are discussed as individual mistakes, not as patterns | → | The same error recurs on the same drug every few months | 1 group |
| Double-checking is signed without a second person actually looking | → | Patients are moved between wards mid-treatment and the chart lags behi | 1 group |
| Handwritten charts are still used when the system is down, which is of | → | Infusion pumps on the ward come from three manufacturers | 1 group |