Quality & Safety

Of the 4 falls recorded this quarter, 3 share three recorded fields: they happened on the general ward between 02:00 and 04:00, the patient had a night sedative administered that shift, and the side-rail was recorded as down. QI-2026-021 (28 Jun), QI-2026-029 (06 Jul) and QI-2026-034 (14 Jul).

The fourth fall — QI-2026-018, OPD waiting area, 11 Jun — shares none of them.

Suggested single CAPA in place of three separate ones: side-rail check recorded at the 02:00 round for every patient with a sedative charted that shift. Owner and due date are yours to set.

Your culture log holds 3 Klebsiella pneumoniae isolates from ICU beds between 06 and 17 Jul, all reported ESBL-positive with the same recorded antibiogram (resistant to ceftriaxone and cefotaxime, sensitive to meropenem). ICU-2 (16 Jul), ICU-5 (11 Jul), ICU-7 (06 Jul). The three months before this held 2 such isolates in total.

Recorded alongside them: hand-hygiene compliance in ICU measured 68.2% in the same window (30 of 44 observed opportunities), and ceftriaxone accounts for 51% of your antimicrobial DDDs.

This is a count of what is in your log, not an outbreak determination. Whether these isolates are related — and whether that relationship is transmission — is a question for the infection control committee, and if needed for typing at a reference lab. HOS has assembled the numbers so that meeting starts with facts instead of a search.

41 reports, 32 intercepted, 9 reached a patient, none with harm above moderate. Prescribing is the largest stage at 14, and 8 of those 14 were a missing frequency or an ambiguous verbal order — both fixable in the prescribing screen rather than by training. Three of the nine that reached a patient were paediatric weight-based dose errors, all in the 4–7 PM pediatric OPD.

Two items need a committee decision, not a CAPA: Losartan / Lisinopril are now stocked adjacently with no control applied, and the 7 unsigned high-alert doses were all on the night shift, which is a staffing question rather than a compliance one.

Everything in her record from admission on 12 Jul to death on 16 Jul, in order, as one document: 4 nursing shift handovers, 26 vitals sets, 31 medication administrations, 5 laboratory reports, 2 chest radiograph reports, 3 consultant notes, 1 rapid-response record from 15 Jul 04:10, the oxygen and NIV charting, and the family communication log. 72 pages. Assembling this by hand takes the quality nurse most of a morning.

Three review steps are outstanding and the committee cannot complete the review without them:

Those three gaps are statements about the documentation, not about the care. What happened, and whether anything should have gone differently, is for the committee to determine and record in its own words.

Four cycles ran; one reached re-audit. Surgical prophylaxis moved 54% → 88% after the OT checklist prompt and the pre-drawn dose, and the remaining 6 misses were all the first procedure of the day — a supply-timing pattern, not a compliance one. Case-sheet completeness measured 71% at baseline and its change (a 24-hour prompt to the admitting doctor) landed on 01 Jul, so a re-audit is possible from August. Transfusion monitoring is 18 of 24 units sampled. Culture-before-antibiotics has a baseline of 44%, which is the same figure the antimicrobial usage card reports independently.

One honest note for the report: three of the four cycles have no re-audit date booked, and a cycle without one tends to become a survey.