Matched clinical audit, 1,860 births, one maternity unit

Midwifery research poster example

A midwifery audit poster has to show four outcomes at once without turning into a table nobody reads. This board plots induction, epidural, breastfeeding and caesarean side by side for two models of care across 1,860 matched Australian births, then prints the counts behind every percentage in the results block.

Create a poster with OneCraftA0 Landscape, printed at 1189 by 841 millimetres

The whole board

The poster at full size, exactly as it prints. Every number, citation and caption on it was written for this example, so the layout is being judged on real content.

KHS
Continuity of midwife care and birth outcomes: a matched audit of 1,860 births at one unit
J. Carmody ¹, P. Nurrung ¹, E. Halvorsen ²
1 Midwifery Research Group, Kelmoor Health Service · 2 Womens and Newborn Unit, Kelmoor Hospital
Abstract
A continuity of midwife care model, where one named midwife and a small group carry a woman through pregnancy, labour and the early weeks, has been offered at this unit since 2024 alongside the standard rostered model. Places in the continuity model are limited, so entry is by availability rather than by risk, which makes a matched comparison possible. This audit covers every birth at the unit over two years. After exclusions, 1,860 births were matched on parity and maternal age, 620 in the continuity model and 1,240 in the standard model at two to one. Four outcomes differ. Induction of labour was 22.1% against 29.0%, epidural use 31.0% against 42.0%, breastfeeding at discharge 91.0% against 84.0% and caesarean birth 24.0% against 27.0%. The largest gap is epidural use, at 11 percentage points. This is an audit of routine records, not a trial, and the limits of that are stated on the board.
Introduction
Continuity models are expensive to roster and hard to staff, so a unit thinking about expanding one wants to know what the current model is already producing before it commits. Published trials of continuity report fewer interventions in labour, but they recruit women who consent to be randomised, and this unit wanted a number from its own population and its own staff. The audit question is deliberately modest. Among women who were clinically similar on the two factors that drive intervention rates hardest at this unit, parity and maternal age, did the women carried by the continuity model have different labour and feeding outcomes from the women carried by the rostered model? The answer is not a causal estimate. It is a description of what the two models produced over two years, with the confounding the audit could not remove named in the discussion rather than buried. The unit also wanted the counts printed beside the percentages, because a rate of 22% reads very differently when it rests on 137 births than when it rests on 14, and the two models here are deliberately unequal in size.
METHODS
Setting. One metropolitan maternity unit, about 1,000 births a year, running a continuity model and a standard rostered model side by side.
Period. Two years of consecutive births, July 2024 to June 2026, taken from the unit birth register and the electronic record.
Matching. Each continuity birth matched to two standard births on parity (first or subsequent) and maternal age band, giving 620 and 1,240.
Outcomes. Induction of labour, epidural in labour, mode of birth and any breastfeeding at discharge, all taken from coded fields, not free text.
Analysis. Crude proportions with exact binomial intervals and risk differences. No adjustment beyond the matching variables was attempted.
PARTICIPANT FLOW
Births at the unit over two years
2,014
Excluded, planned caesarean or birth before 34 weeks
154
Eligible and matched on parity and age
1,860
Continuity of midwife care model
620
Standard rostered model, matched two to one
1,240
Continuity
Standard
Induction
Epidural
Breastfed
Caesarean
0255075100
Figure 1. Induction, epidural, breastfeeding at discharge and caesarean, as a percentage of births. Continuity 620, standard 1,240.
Source: Kelmoor birth register, July 2024 to June 2026.
Results
The counts behind the four percentages are these. Induction of labour: 137 of 620 continuity births, 22.1%, against 360 of 1,240 standard births, 29.0%, a difference of 6.9 points. Epidural in labour: 192 of 620, 31.0%, against 521 of 1,240, 42.0%, a difference of 11.0 points and the largest gap on the board. Any breastfeeding at discharge: 564 of 620, 91.0%, against 1,042 of 1,240, 84.0%, a difference of 7.0 points in the other direction. Caesarean birth: 149 of 620, 24.0%, against 335 of 1,240, 27.0%, a difference of 3.0 points, and the only one of the four whose interval crosses zero. Read together, the three intervention outcomes move in the same direction and the feeding outcome moves with them. The caesarean gap is the smallest and the least certain, which matters because it is the outcome most often quoted from continuity research and the one this audit supports least. Matching held: parity split 41.3% first births in the continuity group against 41.5% standard, and the mean maternal age was 30.4 years in both. Two subgroup readings are worth stating. Among first births, epidural use was 41.7% under continuity and 53.1% under the standard model, a gap of 11.4 points on 256 and 515 births. Among subsequent births it was 23.5% against 34.1%, a gap of 10.6 points on 364 and 725 births, so the epidural difference is not carried by one parity group. Instrumental birth, which the audit collected but did not set out to test, was 13.2% and 15.6%, a difference of 2.4 points.
Discussion
The honest reading is that this is an audit with a selection problem it cannot fix. Places in the continuity model are limited and are taken up by women who book early, and booking early travels with other things that lower intervention rates. Parity and age were matched because they are the two factors this unit can measure reliably in the register, but private health cover, first language, smoking and body mass index were not, and each of those plausibly moves an epidural rate. A trial would settle it and an audit cannot. What the audit can do is set the local size of the difference, which is larger than the unit expected on epidural use and smaller than it expected on caesarean birth. Two operational points follow. The epidural gap is big enough that it would change anaesthetic staffing if the model were expanded, and the caesarean gap is small enough that expansion should not be argued on it. The feeding result is the one most likely to survive a better design, because early postnatal contact is the part of the continuity model that is hardest to substitute and the part women describe most often in the unit feedback.
Key finding
Epidural use was 11 percentage points lower under continuity of midwife care, 31.0% against 42.0%, across 1,860 matched births.
Epidural use 11.0 points lower, 192 of 620 against 521 of 1,240
Induction 6.9 points lower, breastfeeding 7.0 points higher
Caesarean difference 3.0 points, interval crosses zero
Matched on parity and age only, so selection remains
REFERENCES
1.Whitcombe, R. and Ngatai, H. (2025) Continuity models and intervention rates: what audits add to trials. Australian Midwifery Review, 17(2), 96 to 114.
2.Barsoum, L. (2024) Matching on parity and age in maternity audit: a practical limit. Journal of Perinatal Data, 9(3), 211 to 228.
3.Kelmoor Health Service (2026) Womens and Newborn Unit annual birth register report, 2024 to 2026.

Block by block

What each block on the board is for, in the order a reader walks it.

Title band and three authors
The title names the comparison, the birth count and the fact that it is one unit over two years, which is what a reader needs to judge whether the finding travels.
Abstract
How the two models run side by side, why places in the continuity model are allocated by availability rather than risk, the matched design, and all four outcome pairs in one paragraph.
Introduction
Why a unit wants a local number before it expands a model, why published trials answer a slightly different question, and why the counts are printed beside the percentages on this board.
Methods as a labelled protocol
Five entries: the unit and its two models, the two year consecutive period, the two to one match on parity and maternal age band, the four coded outcomes, and the decision not to adjust beyond the matching variables.
Birth flow
A flow block from 2,014 births at the unit, less 154 excluded for a planned caesarean before labour or a birth before 34 weeks, to 1,860 matched births split 620 and 1,240. Every number reconciles.
Figure 1, four outcomes by model
Paired bars for induction, epidural, breastfeeding at discharge and caesarean, one short tick label under each pair and the series legend tucked into the empty top left corner of the plot, clear of the bars. The breastfeeding pair is the only one where continuity sits higher, and it stands out at a glance.
Results and the subgroup reading
The counts behind all four percentages, the matching check on parity and age, and the epidural gap split by parity, 11.4 points among first births and 10.6 points among subsequent births.
Discussion, key message, four points and references
What the audit cannot fix, the callout carrying the 11 point epidural gap, a four item checked list, and three numbered references at the foot of the right column.

What makes this board work

Four outcomes, one figure, two bars each

Induction, epidural, breastfeeding and caesarean are plotted as four paired bars rather than four separate charts, so the whole comparison sits in one figure. Each pair carries a one word tick label, and the caption spells out the full outcome names and the group sizes.

The counts sit under the percentages

The results block gives 192 of 620 against 521 of 1,240 and does the same for the other three outcomes. A percentage on a maternity board means little without its denominator, and the two groups here are deliberately unequal at two to one.

The weakest result is labelled as weak

The caesarean gap of 3.0 points is the outcome most often quoted from continuity research, and this board says its interval crosses zero. Naming the result you support least is what separates an audit poster from a promotional one.

Questions people ask

What goes on a midwifery research poster?

The model of care being compared, the setting and period, how the groups were formed, the outcome definitions taken from coded fields rather than notes, and the counts behind every percentage. For an audit rather than a trial, the selection problem belongs on the board in the discussion, not in a footnote.

How do I show four outcomes without a table?

Plot them as paired bars in one figure and put the counts in the results text. This board gives four categories and two series and prints 137, 192, 564 and 149 against 360, 521, 1,042 and 335 alongside. Keep each category name to a single short word, such as Breastfed, and let the caption carry the full outcome definitions.

Is an audit poster weaker than a trial poster?

It answers a different question. An audit describes what two models produced in one place, which is what a service planning a roster actually needs, and it cannot separate the model from the women who chose it. This board matches on parity and age, names the variables it could not match, and says so plainly.

Do Australian maternity posters need local numbers?

If the point is a local decision, yes. This board is built on one Australian unit of about 1,000 births a year, and its value is that the epidural gap of 11 points is that unit number rather than a pooled figure from elsewhere. State the setting in the title so a reader knows what they are looking at.

Why does the audit exclude planned caesareans before matching?

A planned caesarean is decided before labour, so the woman can never have an induction or a labour epidural and would dilute both rates in either model. The flow removes 154 births for a planned caesarean or birth before 34 weeks, leaving 1,860 to match. Showing that step as its own box lets a reader see what the denominator excludes before reading any percentage.

Build your own in about a minute

The button below opens the generator with this use case already described. Change the wording to match your own, generate, then edit anything you like.

Make one like this

Other poster examples

Veterinary research poster example

A veterinary research poster example should show how animals entered the study, how many finished, and whether the result describes the programme or only the animals that stayed. This A0 landscape board reports a fictional audit of a nurse led weight clinic across four branch practices: owners of 131 overweight cats offered a place, 96 enrolled, 78 completed six months, and a mean fall from 6.4 to 5.7 kilograms in those 78. Three chart figures carry the threshold result, the withdrawals and the monthly curve, and the text keeps completers apart from the programme throughout.

Nutrition research poster example

A dietetics audit poster is read for two things: what the method could actually measure, and which component of the diet is missing. This A0 landscape board photographs 2,400 school lunches across 20 schools, scores each against a guideline, and reports vegetables as the component that fails everywhere.

Genetics research poster example

A genetics poster that reports one variant has to answer two questions before anything else: how many samples, and from where. This board puts both in the title band, then gives the whole width to a hero figure showing the variant frequency rising from 0.08 to 0.21 across five regional cohorts of 1,280 samples each.

Want the steps in the builder? Read Add charts and diagrams, then choose the template, theme and size. For everything this generator can do, see the poster maker.

Sources

Written and checked by the OneCraft team. Last checked .