Article Text

Download PDFPDF

Risk factors and outcomes of management of delayed suprachoroidal haemorrhage following Ahmed glaucoma valve implantation in children
  1. Shantha Balekudaru1,
  2. Tamonash Basu2,
  3. Parveen Sen3,
  4. Pramod Bhende4,
  5. Vijaya Lingam1,
  6. Ronnie George1
  1. 1 Smt Jadhavabhai Nathmal Singhvee Glaucoma Services, Medical Research Foundation, Sankara Nethralaya, Chennai, India
  2. 2 Glaucoma, Medical Research Foundation, Sankara Nethralaya, Kolkata, India
  3. 3 Shri Bhagwan Mahavir Vitreoretinal Services, Medical Research Foundation, Sankara Nethralaya, Chennai, India
  4. 4 Shri Bhagwan Mahavir Vitreoretinal Services, Medical Research Foundation, Chennai, India
  1. Correspondence to Dr Shantha Balekudaru, Smt Jadhavabhai Nathmal Singhvee Glaucoma Services, Medical Research Foundation, Sankara Nethralaya, Chennai 600006, Tamil Nadu, India; shantha.acharya{at}


Aims To assess the incidence, risk factors and outcomes of management of delayed suprachoroidal haemorrhage (DSCH) in children who had undergone Ahmed glaucoma valve implantation.

Methods A retrospective case-control study of eyes which developed DSCH in children <18 years of age who underwent surgery between January 2009 and December 2017 with a follow-up of at least 2 months was performed. Nine cases were compared with 27 age, gender and surgeon matched controls who had undergone surgery during this period.

Results The incidence of DSCH was 4.7% (95% CL 1.5% to 7.7%, 9 eyes of 191 children). There were no significant differences between cases and controls in baseline details except for the number of intraocular pressure (IOP) lowering medications (p=0.01) and follow-up period (p=0.001). Risk factors identified on univariate analysis (p≤0.1) were axial length (p=0.02), diagnosis of primary congenital glaucoma (p=0.05), postoperative hypotony (p=0.07) and aphakia (p=0.1). None of them were found to be significant on multivariate analysis. Five eyes, three with retinal apposition and two with retinal detachment, underwent surgical drainage. There were no significant differences in the outcomes of eyes which underwent drainage compared with those which did not. Failures, defined as IOP>18 mm Hg despite use of medications, loss of light perception, phthisis or removal of the implant were more frequent in cases (three eyes, 33.3%) compared with controls (four eyes, 14.8%) (p=0.002).

Conclusions None of the risk factors analysed in our series proved to be significant. Failures were more common in eyes with choroidal haemorrhage, despite surgical intervention.

  • glaucoma
  • treatment surgery
  • choroid

Statistics from


The use of glaucoma drainage devices in the management of refractory glaucoma in the paediatric population has become increasingly popular. The most commonly reported sight-threatening complications following tube surgery in this age group include retinal detachment, corneal decompensation, suprachoroidal haemorrhage, infection and phthisis.1–3 Suprachoroidal haemorrhage is a devastating complication which may occur either during the surgery or in the early postoperative period.4 While there are numerous publications describing this complication in the adult population, there is a paucity of information available for the paediatric age group.

The risk factors for this complication, reported in adults, include increasing age, systemic hypertension, ischaemic heart disease, the use of anticoagulants, myopia, aphakia, intraoperative as well as postoperative hypotony and prior intraocular surgery including penetrating keratoplasty and pars plana vitrectomy.4–8 A literature search performed on delayed suprachoroidal haemorrhage (DSCH) in children found that the greatest risk was seen following trabeculectomy with Mitomycin C and the use of the Ahmed glaucoma valve in aphakic eyes.9

The aim of our study was to assess the incidence, risk factors and the outcomes of management of eyes which developed DSCH following Ahmed glaucoma valve implantation in children.

Materials and methods

A retrospective case-control study of eyes which had developed DSCH following Ahmed glaucoma valve implantation, performed between January 2009 and February 2017 in children<18 years of age was performed. DSCH was defined as suprachoroidal haemorrhage which occurred in the postoperative period, within a closed system .4 Approval for the study was obtained by the institutional review board. We performed a 1:3 case-control study using a sampling technique reported by Moshfeghi et al.10 Controls that had undergone surgery either 1 month before or 1 month after each case of DSCH were selected. They were age, gender and surgeon matched.

The data retrieved from chart analysis included age, gender, type of glaucoma, the number and type of glaucoma as well as non-glaucoma surgical procedures performed earlier, preoperative intraocular pressure (IOP), number of IOP lowering medications used, preoperative biometric details, postoperative IOP, onset of suprachoroidal haemorrhage, results of ultrasound examination, the duration and details of the management of the suprachoroidal haemorrhage and details of the examination at the last visit including visual acuity (converted to logarithm of minimum angle of resolution), IOP and number of IOP lowering medications.

Surgical technique for AGV implantation

The surgery was performed under general anaesthesia in all patients. The same technique was used in cases as well as controls and remained the same during the entire study period. A similar technique was used by all the surgeons. The Ahmed glaucoma valve (FP7) (New World Medical, Rancho, Cucamonga, California, USA), was implanted in the superotemporal quadrant. A limbal based conjunctival incision was fashioned 4 mm from the limbus. The tenon’s space was dissected posteriorly and the implant was primed and then placed at a distance of about 8–10 mm from the limbus and was then anchored to the surface of the sclera using 9’0’ monofilament nylon sutures. An anterior chamber (AC) maintainer was used in 2 cases and 10 controls. A 23 Gauge needle was used to provide a track of about 2 mm in length for tube implantation, by most of the surgeons, except for one person (RG) who used a 24 G needle (table 1). The tube was trimmed and then inserted using an inserter into the AC parallel to the iris surface, away from the cornea. The tube was ligated in one child and was anchored to the sclera in all the children using 9’0’ monofilament nylon sutures. The exposed portion of the tube was covered with a partial thickness donor corneal graft which was anchored to the sclera using 9’0’ monofilament nylon sutures. The conjunctiva and tenons capsule were opposed in separate layers using 8’0’ polyglactin sutures. All surgeries were performed by surgeons with more than 10 years of surgical experience with drainage devices.

Table 1

Comparison of baseline demographics between cases and controls

Initial management of DSCH included the use of IOP lowering medications if IOP was elevated, and an intensive regimen of topical steroids consisting of 1% prednisolone acetate starting at a dose of 10 times a day and cycloplegic agents such as 2% homatropine bromide or 1% atropine sulfate as well as systemic steroids, that is, tablet prednisolone at a dose of 1 mg per kg body weight, to reduce inflammation. The dosage was tapered over a period of 6–8 weeks, depending on the response to therapy. Serial echographic evaluation with ultrasound was performed to assess progression of clot lysis as well as resolution of the haemorrhage or for the development of an associated retinal detachment.

Surgical technique for drainage of suprachoroidal haemorrhage

All surgeries were performed under general anaesthesia. An infusion line was placed in the AC using an AC maintainer in aphakic eyes. Intraocular saline injection was given to restore the intraocular volume in phakic eyes. In four of the cases, an inferior 180° conjunctival peritomy was performed followed by two radial incisions in the sclera using the microvitreoretinal blade, to reach the suprachoroidal space at the inferotemporal and inferonasal quadrants and the choroidal haemorrhage was drained to the maximum extent possible, through these incisions. The eye volume was restored to normal using intraocular injection of saline and the sclerotomies and conjunctiva were sutured with 7 ‘O’ polyglactin sutures. Pars plana lensectomy with vitrectomy with fluid gas exchange, endolaser photocoagulation and silicon oil injection was performed in one case in which the suprachoroidal haemorrhage was associated with a rhegmatogenous retinal detachment.

Definition of outcomes


IOP≥6 mm Hg and ≤18 mm Hg with or without IOP lowering medications.


IOP>18 mm Hg despite use of IOP lowering therapy, loss of vision following surgery, that is, <2.3 (LogMAR), phthisis or removal of the implant.

Statistical methods

SPSS V.14 was used for the statistical analysis. The independent-samples Student ‘t’ test was used for comparison of normally distributed data. Non-parametric tests such as the Mann Whitney ‘U’ test and Wilcoxon Signed Rank test were used for data which was not normally distributed. Categorical data were analysed with the χ² test and Fisher’s exact test. Statistical significance was set at p<0.05. Risk factors for the development of suprachoroidal haemorrhage were analysed by logistic regression analysis. Risk factors with a significance of p≤0.1 on univariate analysis were included in a multivariate analysis.


The baseline demographics are depicted in table 1. The cases and controls were matched for surgeons (p=0.32) as well as for age and gender.

There were no significant differences between cases and controls in terms of the baseline demographics except in the number of IOP lowering medications and the follow-up period (table 2). Special precautions to reduce the risk of a choroidal haemorrhage included use of an AC maintainer in two cases and tube ligation in one case and use of an AC maintainer in nine and prophylactic sclerotomy in one, of the controls, respectively (p=0.84).

Table 2

Comparison of preoperative data

One hundred and ninety-one children less than 18 years of age had undergone Ahmed glaucoma valve implantation during the study period. Nine cases developed DSCH (4.7%, 95% CL 1.5% to 7.7%) following surgery. Seven patients presented with suprachoroidal haemorrhage on the first postoperative day; one patient on the fourth day and one on the seventh day following surgery. The mean duration of choroidal haemorrhage from the onset until complete resolution or development of phthisis was 61.8±26.9 days. Ultrasound examination revealed choroidal haemorrhage in all nine eyes with central retinal apposition in 3 (figure 1) of these cases and choroidal haemorrhage with central retinal apposition, associated with retinal detachment in two eyes.

Figure 1

Ultrasound B scan of the left eye of a 10-year-old child who developed DSCH on the first postoperative day after AGM. B scan shows 360° choroidal detachment with suprachoroidal echoes suggestive of haemorrhage. Few echoes are also seen in the vitreous cavity suggestive of associated vitreous haemorrhage. AGM, anti-glaucoma medication; DSCH, delayed suprachoroidal haemorrhage.

Surgical drainage of the choroidal detachment was performed in five eyes. The mean duration between the onset of the choroidal haemorrhage and surgical drainage was 13.6±9.8 days (5–30). The indications for surgical drainage included retinochoroidal adhesion with central retinal apposition in three eyes and retinal detachment associated with choroidal haemorrhage in two eyes. The comparison of the preoperative details and postoperative outcomes between eyes which underwent surgical drainage and those which did not is depicted in table 3. There were no significant differences in either the preoperative details or in the postoperative outcomes between those who underwent surgical drainage and those who did not except for the fact that none of the eyes which were conservatively managed required medical therapy for IOP control.

Table 3

Comparison of baseline demographics and postoperative outcomes among eyes which underwent surgical drainage of choroidal haemorrhage and those who did not

Postoperative outcomes are depicted in table 4. The visual acuity at the end of 2 months was worse in cases compared with controls. There were more failures among those who developed choroidal haemorrhage (six eyes, 66%), than among the controls (four eyes, 14.8%), p=0.002. The reasons for failure included loss of vision following surgery, that is, <2.3 (LogMAR), phthisis and removal of the implant among cases. The reasons for failure in the control group included elevated IOP in two eyes, Ahmed Glaucoma valve (AGV) removal in one eye and due to loss of vision in one eye.

Table 4

Comparison of postoperative outcomes

Risk factors assessed for development of choroidal haemorrhage included age, gender, preoperative glaucoma diagnosis, baseline IOP, preoperative axial length, pachymetry, lens status, pars plana vitrectomy performed along with AGV implantation and postoperative hypotony (IOP<6 mm Hg) within the first postoperative week (table 5). Increasing axial length, horizontal corneal diameter, diagnosis of primary congenital glaucoma and postoperative hypotony were found to be risk factors (p≤0.1) on univariate analysis. None of these factors remained significant on multivariate analysis although axial length reached borderline significance (p=0.07).

Table 5

Risk factors for choroidal haemorrhage


The incidence of suprachoroidal haemorrhage in adults varies from 1.6% following filtering surgery, up to 6% in those undergoing surgery with glaucoma drainage devices.4 6–8

The incidence following the use of glaucoma drainage devices in children varies from 3.2% to 4.75%.11 12 There is a paucity of literature on the risk factors and management of this complication in the paediatric age group. In a literature review conducted for the period extending from 1966 to 2008, delayed suprachoroidal haemorrhage was identified in nine cases among five case series: three following trabeculectomy with Mitomycin C and six following Ahmed valve implantation of which adjunctive Mitomycin-C was used in two cases. Four of the eyes which had undergone AGV were definitely aphakic and two of the eyes which underwent AGV and the three eyes which underwent trabeculectomy, were possibly aphakic. The authors attributed increased risk of DSCH following AGV implantation, in aphakic eyes, with the use of adjunctive Mitomycin C.9

Choroidal haemorrhage occurs more frequently following the use of glaucoma drainage devices compared with filtering surgeries such as trabeculectomy. Tuli et al reported a higher incidence of DSCH after tube shunts compared with trabeculectomy in adults (p<0.001). The significant risk factors in their series included white race (p=0.012), anticoagulation (p=0.034), postoperative hypotony (p=0.033) and aphakia/ AC intraocular lens implantation (p=0.002).5 Jeganathan et al reported an increased risk after drainage implantation as well; OR 3.4, 95% CL 1.9 to 5.4, p<0.001. The risk factors which they described included postoperative hypotony (p<0.001, aphakia (p<0.001), prior intraocular surgery (p=0.002, hypertension, (p<0.001), anticoagulation (p<0.002), ischaemic heart disease (p=0.001) and respiratory disease (p=0.008).6 Vaziri et al, in a retrospective study using the national insurance claim based databases from 2007 to 2011 noted that the cumulative incidence of suprachoroidal haemorrhage occurring within 3 months of surgery was twice as likely to occur following shunt surgery than after trabeculectomy.7

Rupture of the short and long posterior ciliary arteries traversing the suprachoroidal space is said to occur when this space suddenly expands due to the acute onset of a suprachoroidal haemorrhage, precipitated by postoperative hypotony.13 Early postoperative hypotony was a significant risk factor on univariate analysis, but not in the multivariate analysis in our series. Increasing axial length was found to be of borderline significance in our series. This factor was reported as a risk factor by Reibaldi et al in their series on the development of DSCH in vitrectomised eyes (OR 2.57 [95% CL 1.42 to 7.52] [p<0.001]).8 This could be associated with reduced ocular rigidity, which may have predisposed these eyes to postoperative hypotony, thereby precipitating the onset of DSCH. Myopia, aphakia and vitrectomy, all of which are associated with lower ocular rigidity which in turn may be associated with increased choroidal vascular fragility, have been found to increase the risk of suprachoroidal haemorrhage.4

Measures to reduce the risk of choroidal haemorrhage include control of IOP to the maximum extent possible with medical therapy, discontinuing anticoagulants, if in use, control of hypertension and prevention of intraoperative and postoperative hypotony.4 14 The preoperative IOP was similar in cases and controls in our series, although the number of IOP lowering agents was higher in the control group. Measures to prevent intraoperative hypotony in our series were not universally applied; two cases developed DSCH despite the use of an AC maintainer in two cases and ligation of the tube in one case. Other prophylactic measures described include gradual decompression of the globe during surgery and avoiding valsalva manoeuvres in the postoperative period, since postoperative emesis is one of the risk factors associated with development of DSCH.

The goals of surgical management of DSCH in our series were to drain the choroidal haemorrhage, reattach the retina and maintain IOP control. This was achieved in five eyes (55.9%) in our series. The remainder developed phthisis. The timing of surgical drainage remains controversial. Clot lysis is said to occur within 7–14 days and is said to facilitate surgical drainage of the haemorrhage.15 In the event of elevated IOP, inability to close the surgical wound and associated retinal detachment, immediate surgery is advocated. Pakhravan et al reported the results of surgical drainage, performed through inferior sclerotomies within 36 hours of surgery in seven eyes with suprachoroidal haemorrhage with favourable outcomes in terms of IOP control and improvement of visual acuity.16 None of these eyes had concomitant retinal detachment in contrast to our series in which two eyes had developed this complication. Jin et al performed surgical drainage within 48 hours in eight eyes with suprachoroidal haemorrhage, of which two had undergone trabeculectomy.17 Five eyes underwent pars plana vitrectomy with silicon oil tamponade and three eyes underwent drainage through a sclerotomy. IOP control and visual acuity improved significantly in their series (p<0.001). All of these patients were adults however.

Although three of five patients in our series underwent surgical drainage within 1 week of the onset of choroidal haemorrhage, the duration of the haemorrhage prior to resolution was prolonged (10 weeks in those who underwent drainage compared with 7 weeks in those who did not). In addition, appositional choroidal haemorrhage occurred in 5 eyes (55.5%) in our series which may have contributed to the poor visual outcome. Moshfeghi et al reported their outcomes of appositional suprachoroidal haemorrhage in 37 cases.10 One of the factors associated with poor visual outcome in their series included prolonged apposition of >30 days (p=0.01). They also reported rhegmatogenous retinal detachment in 14 eyes (38%), hypotony in 9 eyes (24%) phthisis in 8 eyes (22%) and tractional retinal detachment in 5 eyes (14%). Poor visual outcomes were reported by Tuli et al in their cases, which was significantly worse than in controls (p<0.009) in their series.6 Poor visual outcomes were reported by Jeganathan et al as well, compared with controls in their series (p=0.002), due to the presence of tractional retinal detachment in three eyes, hypotony maculopathy in two eyes and phthisis in three eyes.7

The merits of our study include the following factors: surgery performed at a single, tertiary care centre, with a single type of implant using the same surgical technique, performed by experienced surgeons. All the demerits of a retrospective study with inherent biases apply to this study as well. Prophylactic measures to prevent this complication were not applied uniformly, in either the cases or the controls. Details of events such as emesis which is common following general anaesthesia were not available in the majority of our case files. Moreover, DSCH is a rare complication and therefore difficult to assess in a prospective manner.


The incidence of choroidal haemorrhage in our series was about 4.7%. None of the risk factors assessed for the development of suprachoroidal haemorrhage were found to be significant. Success rates and visual outcome was poor in eyes with choroidal haemorrhage in our series, despite surgical intervention.


Natarajan Viswanathan for statistical assistance.



  • Contributors TB: data collection, PS: contributed to surgical management of the choroidal haemorrhage and reviewed manuscript, PB: contributed to surgical management of the choroidal haemorrhage and reviewed manuscript, VL and RG: contributed to the data used and reviewed manuscript.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Ethics approval Institutional Review Board (Ethics Committee) of Vision Research Foundation, Sankara Nethralaya, Chennai, India. Approval Number: 742-2018-P, dt. 20.12.2018.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement All data relevant to the study are included in the article or uploaded as supplementary information.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • At a glance
    Keith Barton Jost B Jonas James Chodosh