> ## Content Index
> Fetch the complete content index at: https://www.swisseyemd.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Before the Examination, and After the Discharge
- URL: https://www.swisseyemd.com/before-the-examination-and-after-the-discharge/
- Published: 2026-10-14T06:00:41.000Z
- Updated: 2026-10-14T13:36:14.000Z
- Description: Two questions the evidence has not settled: whether an acute PVD can be stratified before anyone has looked at the retina, and whether the six-week discharge is safe.
- Author: Miguel Castilla
- Tags: The Six-Week Rule, Acute PVD

0:00 

/1:38 

1× 

The last two entries looked at the same problem from two sides. The rates we quote vary by an order of magnitude for reasons of catchment and of citation, and the symptom we ask about most confidently is the one the evidence cannot agree on. That leaves two questions: one about the telephone call, before she has been examined, and one about the discharge, after.

The first is whether any of this can be done before the patient is in front of us.

Most of the findings that move the probability require a dilated pupil: vitreous haemorrhage, vitreous pigment, lattice degeneration. Over the telephone, two things that matter can still be asked for. One is the history of the fellow eye. The other is the one symptom with a consistent signal, a loss of vision the patient has noticed, which the [last entry](https://swisseyemd.com/what-a-flash-is-worth/?ref=swisseyemd.com) showed takes a pre-test probability of 14% to 45%. Yet the decision that determines who is examined today and who is examined in a fortnight is made before anything else is known, by whoever answers the telephone.

One group has tried to move the decision upstream. Over four months in 2020, the triage staff of a single academic centre in Michigan put a structured questionnaire of twenty-three questions, with seventy pre-specified answers, to callers whose symptoms suggested a posterior vitreous detachment. The 193 who were examined within six weeks make up the cohort, and fourteen of them had a retinal tear or a detachment. From those data the authors built a clinical risk score using seven of the twenty-three questions and fifteen of the seventy possible answers, and reported an area under the curve above 0.90 for both the regression model and the score; the multivariable model's figure is 0.9205\. A score of 20 captured every patient who turned out to have a tear or a detachment while excluding nearly half of those who did not.

What carried the weight in that model is worth a look, because it is the signal the last entry found. The strongest term was constant blurred vision or a curtain or veil in the field, at an odds ratio of 12.01, confidence interval 2.83 to 50.99\. Calling within twenty-four hours of onset, compared with calling after more than seventy-two, reached an odds ratio of 3.01 with a confidence interval from 0.64 to 14.20, which does not exclude no effect at all. In the 8,305 patients of the Kaiser Permanente cohort, blurred vision at triage carried an odds ratio of 2.67, and flashes 0.76\. Three datasets built in different ways point the same way: what the patient can no longer see says more than what the patient sees.

**That telephone score is the most direct published attempt to stratify an acute posterior vitreous detachment before anyone has looked at the retina, and it rests on fourteen events at one centre.** The authors say so themselves, in the limitations: the study lacks independent validation, they are prospectively following a second set of patients, and the work was carried out at a single academic medical centre whose patient base may differ from other settings. A model with seven predictors fitted to fourteen events is a hypothesis with a confidence interval, and a discrimination statistic computed on the data that produced the model is not a performance estimate. None of that makes it wrong. It makes it unvalidated, and the distinction matters more here than it usually does, because what a triage score gets wrong is who does not get seen.

The second question sits at the other end. Once she has been examined, indented, and found to have nothing, is it safe to let her go at six weeks?

The six-week convention exists because retinal breaks were thought to declare themselves inside it. Two cohorts have now followed patients well past it. In 7,999 eyes seen by an academic retina practice, 44.5% of the delayed breaks were found more than six weeks after presentation, and among the delayed detachments the figure was 67.5%. In a separate cohort of 389 eyes followed for up to six years, 55% of the delayed tears appeared after six weeks, and the survival curve did not reach half of them until 4.63 months.

Which would matter less if the patients told us. In that second cohort, seventy-five per cent of the patients who developed a delayed tear had no new or increased symptoms when it was found. In the first, 40.2% of the delayed breaks came with new or worsening symptoms. Detachments behave differently: more than two-thirds of the delayed detachments in the first cohort were symptomatic when found, and 57.5% were macula-off. The safety-netting advice that carries most of the weight in current practice, the instruction to come back if anything changes, is being asked to catch tears that mostly announce nothing, and detachments that do announce themselves, more than half of them once the macula has gone.

![A horizontal timeline of one clinical episode with two vertical markers. The first, the telephone call, has a panel listing reported symptoms and past history. The second, discharge at six weeks, has a panel listing the findings an indented examination provides, including vitreous haemorrhage, vitreous pigment and lattice degeneration. A third panel to the right lists what arrives after discharge: 44.5 and 55 per cent of delayed breaks, 67.5 per cent of delayed detachments, and 75 per cent of delayed breaks with no new symptoms.](https://storage.ghost.io/c/0e/7b/0e7b0c0e-ac22-42d2-aef9-0c07388f00c5/content/images/2026/09/Gemini_Generated_Image_83dk5g83dk5g83dk.jpg)

Figure 3\. The two moments at which the interval is decided, and what is known at each. At the first, only what the patient reports. At the second, everything the examination provides, and the episode is closed. The panel on the right is what arrives after the second marker. Data: Uhr JH, et al. Ophthalmology 2020;127:516-22; Jindachomthong KK, et al. Ophthalmol Retina 2023;7:318-24.

**The one comparative test of that advice among the papers behind this series found no signal and no events.** Fifty patients counselled by a nurse practitioner, with a leaflet, were compared with fifty historical controls who were not; returns with no change in symptoms fell from three to none, at a one-tailed Fisher exact P of 0.13, which the authors themselves described as not highly significant. Not one patient in either group re-presented with a retinal break, a detachment or a vitreous haemorrhage. A hundred patients and zero events cannot tell us whether structured advice detects anything. It can only tell us it did not obviously cost anything.

So the position is this. Before the examination, one unvalidated score derived on fourteen events. After the discharge, roughly half of the delayed tears and two-thirds of the delayed detachments falling outside the window, the tears mostly without warning. And in between, a guideline whose follow-up table stratifies on the presence of blood.

Both questions have the same shape. They ask whether the interval can be matched to the patient rather than to the average of a literature, and at the two moments where matching it would change what happens. Whether that can be done with the evidence that exists, and what a structured attempt at it would look like, is the subject of poster F.127 at the 29th EVER Congress in Florence on 16 October. The twenty-six papers behind it are already on a reference page, [swisseyemd.com/pvd](https://swisseyemd.com/pvd?ref=swisseyemd.com).

Until then, the question the last two entries have been building towards. If the findings that stratify her are only available after we have examined her, and half the events arrive after we have discharged her, at which moment in this episode are we actually deciding anything?

---

**References**

1. Hollands H, Johnson D, Brox AC, Almeida D, Simel DL, Sharma S. Acute-onset floaters and flashes: is this patient at risk for retinal detachment? JAMA. 2009;302(20):2243-9\. doi:10.1001/jama.2009.1714.
2. Balikov DA, Zhou Y, Miller JML. A telephone triage system for patients calling with symptoms of a posterior vitreous detachment. Ophthalmol Retina. 2023;7(6):516-26\. doi:10.1016/j.oret.2023.01.003.
3. Seider MI, Conell C, Melles RB. Complications of acute posterior vitreous detachment. Ophthalmology. 2022;129(1):67-72\. doi:10.1016/j.ophtha.2021.07.020.
4. Uhr JH, Obeid A, Wibbelsman TD, et al. Delayed retinal breaks and detachments after acute posterior vitreous detachment. Ophthalmology. 2020;127(4):516-22\. doi:10.1016/j.ophtha.2019.10.020.
5. Jindachomthong KK, Cabral H, Subramanian ML, et al. Incidence and risk factors for delayed retinal tears after an acute, symptomatic posterior vitreous detachment. Ophthalmol Retina. 2023;7(4):318-24\. doi:10.1016/j.oret.2022.10.012.
6. Singh AJ, Seemongal-Dass RR. The influence of counselling on patient return following uncomplicated posterior vitreous detachment. Eye (Lond). 2001;15(Pt 2):152-4\. doi:10.1038/eye.2001.50.
7. Kim SJ, Bailey ST, Kovach JL, et al. Posterior Vitreous Detachment, Retinal Breaks, and Lattice Degeneration Preferred Practice Pattern. Ophthalmology. 2025;132(4):P163-P196\. doi:10.1016/j.ophtha.2024.12.023.