Interventions for dissociated vertical deviation

  • Review
  • Intervention

Authors


Abstract

Background

The term "strabismus" describes misalignment of the eyes. One or both eyes may deviate inward, outward, upward, or downward. Dissociated vertical deviation (DVD) is a well-recognized type of upward drifting of one or both eyes, which can occur in children or adults. DVD often develops in the context of infantile- or childhood-onset horizontal strabismus, either esotropia (inward-turning) or exotropia (outward-turning). For some individuals, DVD remains controlled and can only be detected during clinical testing. For others, DVD becomes spontaneously "manifest" and the eye drifts up of its own accord. Spontaneously manifest DVD can be difficult to control and often causes psychosocial concerns. Traditionally, DVD has been thought to be asymptomatic, although some individuals have double vision. More recently it has been suggested that individuals with DVD may also suffer from eyestrain. Treatment for DVD may be sought either due to psychosocial concerns or because of these symptoms. The standard treatment for DVD is a surgical procedure; non-surgical treatments are offered less commonly. Although there are many studies evaluating different management options for the correction of DVD, a lack of clarity remains regarding which treatments are most effective.

Objectives

The objective of this review was to determine the effectiveness and safety of various surgical and non-surgical interventions in randomized controlled trials of participants with DVD.

Search methods

We searched CENTRAL (which contains the Cochrane Eyes and Vision Trials Register) (2015, Issue 8), Ovid MEDLINE, Ovid MEDLINE In-Process and Other Non-Indexed Citations, Ovid MEDLINE Daily, Ovid OLDMEDLINE (January 1946 to August 2015), EMBASE (January 1980 to August 2015), PubMed (1948 to August 2015), Latin American and Caribbean Health Sciences Literature Database (LILACS) (1982 to August 2015), the metaRegister of Controlled Trials (mRCT) (www.controlled-trials.com) (last searched 3 February 2014), ClinicalTrials.gov (www.clinicaltrials.gov), and the WHO International Clinical Trials Registry Platform (ICTRP) (www.who.int/ictrp/search/en). We did not use any date or language restrictions in the electronic searches for trials. We last searched the electronic databases on 3 August 2015.

Selection criteria

We included randomized controlled trials (RCTs) of surgical and non-surgical interventions for the correction of DVD.

Data collection and analysis

We used standard procedures expected by Cochrane. Two review authors independently completed eligibility screening, data abstraction, 'Risk of bias' assessment, and grading of the evidence.

Main results

We found four RCTs eligible for inclusion in this review (248 eyes of 151 participants between the ages of 6 months to 22 years). All trials were assessed as having unclear risk of bias overall due to insufficient reporting of study methods. One trial was conducted in Canada and compared anteriorization of the inferior oblique muscle with resection versus anteriorization of the inferior oblique muscle alone; one in the USA compared superior rectus recession with posterior fixation suture versus superior rectus recession alone; and two in the Czech Republic compared anteriorization of the inferior oblique muscle versus myectomy of the inferior oblique muscle.

Only one trial reported data that allowed analysis of the primary outcome for this review, the proportion of participants with treatment success. The difference between inferior oblique anteriorization plus resection versus inferior oblique anteriorization alone was uncertain when measured at least four months postoperatively (risk ratio 1.13, 95% confidence interval 0.60 to 2.11, 30 participants, very low-quality evidence). Three trials measured the magnitude of hyperdeviation, but did not provide sufficient data for analysis. All four trials reported a relatively low rate of adverse events; hypotropia, limited elevation, and need for repeat surgery were reported as adverse events associated with some of the surgical interventions. No trials reported any other secondary outcome specified for our review.

Authors' conclusions

The four trials included in this review assessed the effectiveness of five different surgical procedures for the treatment of DVD. Nevertheless, insufficient reporting of study methods and data led to methodological concerns that undermine the conclusions of all studies. There is a pressing need for carefully executed RCTs of treatment for DVD in order to improve the evidence for the optimal management of this condition.

Resumen

Intervenciones para la desviación vertical disociada

Antecedentes

El término "estrabismo" describe la alineación incorrecta de los ojos. Uno o ambos ojos pueden desviarse hacia adentro, hacia afuera, hacia arriba o hacia abajo. La desviación vertical disociada (DVD) es un tipo bien reconocido de desviación hacia arriba de uno o ambos ojos que puede ocurrir en niños o adultos. La DVD a menudo se desarrolla en el contexto del estrabismo horizontal infantil o de aparición en la infancia, o esotropía (voltear hacia adentro) o exotropía (voltear hacia afuera). En algunos individuos, la DVD se mantiene controlada y solamente se puede detectar durante una prueba clínica. En otros, la DVD está espontáneamente "manifiesta" y el ojo se desvía hacia arriba por sí solo. La DVD espontáneamente manifiesta puede ser difícil de controlar y a menudo causa problemas psicosociales. Tradicionalmente se ha pensado que la DVD es asintomática, aunque algunos individuos presentan diplopía. Más recientemente se ha indicado que los individuos con DVD también pueden presentar fatiga visual. El tratamiento para la DVD se puede solicitar debido a problemas psicosociales o a estos síntomas. El tratamiento estándar para la DVD es un procedimiento quirúrgico; los tratamientos no quirúrgicos se ofrecen con menos frecuencia. Aunque hay muchos estudios que evalúan diferentes opciones de tratamiento para la corrección de la DVD, aún no está claro qué tratamientos son los más eficaces.

Objetivos

El objetivo de esta revisión fue determinar la efectividad y la seguridad de diversas intervenciones quirúrgicas y no quirúrgicas en ensayos controlados aleatorios de participantes con DVD.

Métodos de búsqueda

Se realizaron búsquedas en CENTRAL (que contiene el registro de ensayos del Grupo Cochrane de Trastornos de los Ojos y la Visión [Cochrane Eyes and Vision Group]) (2015, número 8), Ovid MEDLINE, Ovid MEDLINE In-Process and Other Non-Indexed Citations, Ovid MEDLINE Daily, Ovid OLDMEDLINE (desde enero de 1946 hasta agosto de 2015), EMBASE (desde enero de 1980 hasta agosto de 2015), PubMed (desde enero de 1948 hasta agosto de 2015), Latin American and Caribbean Health Sciences Literature Database (LILACS) (desde enero de 1982 hasta agosto de 2015), en el metaRegister of Controlled Trials (mRCT) (www.controlled-trials.com) (última búsqueda el 3 de febrero de 2014), ClinicalTrials.gov (www.clinicaltrials.gov) y en la Plataforma de registros internacionales de ensayos clínicos (ICTRP) de la OMS (www.who.int/ictrp/search/en). No se aplicó ninguna restricción de fecha o idioma en las búsquedas electrónicas de ensayos. Se buscó por última vez en las bases de datos electrónicas el 3 de agosto de 2015.

Criterios de selección

Se incluyeron los ensayos controlados aleatorios (ECA) de intervenciones quirúrgicas y no quirúrgicas para la corrección de la DVD.

Obtención y análisis de los datos

Se utilizaron los procedimientos metodológicos estándar previstos por la Colaboración Cochrane. Dos revisores de forma independiente completaron el examen de la elegibilidad, el resumen de los datos, la evaluación del "riesgo de sesgo" y la clasificación de las pruebas.

Resultados principales

Se encontraron cuatro ECA elegibles para inclusión en esta revisión (248 ojos de 151 participantes con edades entre seis meses y 22 años). Todos los ensayos se evaluaron como riesgo de sesgo general incierto debido al informe insuficiente de los métodos de estudio. Un ensayo se realizó en Canadá y comparó la anteriorización del músculo oblicuo inferior con resección versus la anteriorización del músculo oblicuo inferior sola; un estudio en los EE.UU. comparó la recesión del recto superior con fijación posterior con sutura versus la recesión del recto superior sola; y dos en la República Checa compararon anteriorización del músculo oblicuo inferior versus miectomía del músculo oblicuo inferior.

Solamente un ensayo informó datos que permitieron el análisis del resultado primario para esta revisión, proporción de participantes con éxito del tratamiento. La diferencia entre la anteriorización del oblicuo inferior más resección versus la anteriorización del oblicuo inferior sola fue incierta cuando se midió al menos cuatro meses después de la cirugía (cociente de riesgos 1,13; intervalo de confianza del 95%: 0,60 a 2,11; 30 participantes, pruebas de muy baja calidad). Tres ensayos midieron la magnitud de la hiperdesviación, pero no proporcionaron suficientes datos para el análisis. Los cuatro ensayos informaron una tasa relativamente baja de eventos adversos; la hipotropía, la elevación limitada y la necesidad de cirugía repetida se informaron como eventos adversos asociados con algunas de las intervenciones quirúrgicas. Ningún ensayo informó otro resultado secundario especificado en la revisión.

Conclusiones de los autores

Los cuatro ensayos incluidos en esta revisión evaluaron la efectividad de cinco procedimientos quirúrgicos diferentes para el tratamiento de la DVD. No obstante, el informe insuficiente de los métodos y los datos de los estudios provocaron inquietudes metodológicas que debilitan las conclusiones de todos los estudios. Hay una necesidad apremiante de ECA realizados de manera cuidadosa del tratamiento de la DVD con el objetivo de mejorar las pruebas para el tratamiento óptimo de esta afección.

Plain language summary

Treatment for eyes that drift upwards

Review question
The aim of this review was to evaluate the effectiveness of surgical and non-surgical treatments for dissociated vertical deviation.

Background
Eye misalignment (strabismus) is the drifting of one or both eyes, which can be inward, outward, upward, or downward. This review evaluated the treatment for a specific type of upward drifting of one or both eyes known as dissociated vertical deviation (DVD). DVD can occur in both children and adults. For some people, DVD is controlled and is only detectable during testing. In others, DVD happens all of a sudden as the eye drifts up of its accord. It can be hard for the person to gain control of the eye, which can cause distress to the person in social situations. The condition also may cause double vision or eyestrain.

Surgery is the common treatment for DVD. Treatments that do not involve surgery are uncommon. There is limited evidence about the effectiveness of treatments (either surgical or non-surgical) for DVD.

Study characteristics
We conducted the search for studies on 3 August 2015. We found four randomized controlled trials (RCTs) of surgical treatment for DVD. We found no studies evaluating non-surgical treatments. One trial was conducted in Canada and compared a surgical repositioning procedure (anteriorization of the inferior oblique muscle) with or without resection; one in the USA compared surgical weakening of an eye muscle (superior rectus recession) with or without augmentation with a fixation suture; and two in the Czech Republic compared anteriorization of the inferior oblique muscle versus removal of a piece of the inferior oblique muscle (myectomy).

Key results
Only one of the RCTs examined what we wanted to know: the proportion of participants who had surgical success. There was insufficient information available to determine the differences between any of the surgical procedures with respect to surgical success or any other outcome relevant to our review. The most common adverse events from the surgical procedures were downward drifting of the eye after surgery (hypotropia), limited upward movement of the eye, and need for repeat surgery.

Quality of the evidence
All four of the included studies had flaws in design, execution, or both that weaken their conclusions. There is a need for well-designed, rigorously conducted RCTs of treatments for DVD to provide more reliable evidence for the management of this condition.

Streszczenie prostym językiem

Leczenie zeza pionowego ku górze

Pytanie badawcze
Celem tego przeglądu była ocena skuteczności chirurgicznego i niechirurgicznego leczenia rozkojarzonego odchylenia pionowego.

Wprowadzenie
Odchylenie oka (zez) polega na mimowolnym ustawieniu jednej lub obu gałek ocznych, które może być do wewnątrz, na zewnątrz, ku górze lub ku dołowi. W niniejszym przeglądzie poddano ocenie leczenie konkretnego typu zeza, w którym jedna lub obie gałki oczne mimowolnie odchylone są ku górze, określanego jako zdysocjowane odchylenie pionowe (DVD). DVD może występować zarówno u dzieci, jak i u dorosłych. Niektóre osoby potrafią kontrolować DVD i wykrywa się je tylko podczas badania. U innych, DVD zdarza się nagle jako mimowolne odchylenie gałki ocznej ku górze. Utrzymanie kontroli nad ustawieniem oka może stanowić trudność i powodować niepokój w sytuacjach społecznych. Stan ten może również powodować dwojenie lub pogorszenie ostrości wzroku.

Powszechną metodą leczeniem DVD jest zabieg operacyjny. Leczenie nieobejmujące zabiegów chirurgicznych jest niezbyt częste. Dane naukowe na temat skuteczności leczenia (zarówno chirurgicznego, jak i niechirurgicznego) w przypadku DVD są ograniczone.

Charakterystyka badań
Wyszukiwanie badań przeprowadziliśmy 3 sierpnia 2015 r. Odnaleźliśmy cztery badania kliniczne z randomizacją (RCT) na temat chirurgicznego leczenie DVD. Nie odnaleźliśmy natomiast żadnych badań oceniających metody niechirurgiczne. Jedno badanie zostało przeprowadzone w Kanadzie i porównywano w nim procedurę chirurgicznego przemieszczenia mięśnia (przesunięcia przyczepu mięśnia skośnego dolnego do przodu) z jego resekcją lub bez resekcji; jedno w USA, w którym porównywano chirurgiczne cofnięcie mięśnia oka (recesja mięśnia prostego górnego) ze wzmocnieniem lub bez wzmocnienia mięśnia przy pomocy szwów; oraz dwa w Czechach porównujące przesuniecie przyczepu mięśnia skośnego dolnego do przodu w odniesieniu do usunięcia części mięśnia skośnego dolnego (miektomia).

Główne wyniki
Tylko w jednym badaniu RCT oceniano to, co chcieliśmy wiedzieć: odsetek uczestników, u których operacja okazała się skuteczna. Nie było wystarczających informacji, aby określić różnice pomiędzy którąkolwiek z procedur chirurgicznych w odniesieniu do powodzenia operacji lub jakichkolwiek innych wyników badania istotnych dla niniejszego przeglądu. Najczęstszymi działaniami niepożądanymi w przypadku procedur chirurgicznych było mimowolne odchylenie oka ku dołowi po zabiegu (hypotropia), ograniczenie ruchomości oka ku górze oraz konieczność przeprowadzenia powtórnej operacji.

Jakość danych naukowych
Wszystkie cztery spośród włączonych badań miały wady w zakresie projektowania badania, jego wykonania lub w obu obszarach, w konsekwencji osłabiając ich wnioski. Należałoby w przyszłości stworzyć dobrze zaprojektowane, rygorystycznie przeprowadzone badania RCT dotyczące leczenia DVD, aby zapewnić bardziej wiarygodne dane naukowe na temat postępowania w tej chorobie.

Uwagi do tłumaczenia

Tłumaczenie: Dawid Storman Redakcja: Agnieszka Ole, Piotr Kanclerz

Resumen en términos sencillos

Tratamiento para los ojos que se desvían hacia arriba

Pregunta de la revisión
El objetivo de esta revisión fue evaluar la efectividad de los tratamientos quirúrgicos y no quirúrgicos para la desviación vertical disociada.

Antecedentes
La alineación incorrecta de los ojos (estrabismo) es la desviación de uno o ambos ojos, que puede ser hacia adentro, hacia afuera, hacia arriba o hacia abajo. Esta revisión evaluó el tratamiento de un tipo específico de desviación hacia arriba de uno o ambos ojos, conocida como desviación vertical disociada (DVD). La DVD puede ocurrir en niños y en adultos. En algunos pacientes la DVD se controla y solamente es detectable durante las pruebas. En otros, la DVD aparece de repente cuando el ojo se desvía espontáneamente. Puede ser difícil para el paciente controlar el ojo, lo que le puede causar angustia en situaciones sociales. La afección también puede causar diplopía o fatiga visual.

La cirugía es el tratamiento habitual para la DVD. Los tratamientos que no incluyen cirugía son poco frecuentes. Hay pruebas limitadas acerca de la efectividad de los tratamientos (quirúrgicos o no quirúrgicos) para la DVD.

Características de los estudios
La búsqueda de los estudios se realizó el 3 agosto de 2015. Se encontraron cuatro ensayos controlados aleatorios (ECA) de tratamiento quirúrgico para la DVD. No se encontraron estudios que evaluaran tratamientos no quirúrgicos. Un ensayo se realizó en Canadá y comparó un procedimiento de reubicación quirúrgica (anteriorización del músculo oblicuo inferior) con o sin resección; uno en los EE.UU. comparó el debilitamiento quirúrgico de un músculo del ojo (recesión del recto superior) con o sin aumento con fijación con sutura; y dos en la República Checa compararon anteriorización del músculo oblicuo inferior versus extracción de una parte del músculo oblicuo inferior (miectomía).

Resultados clave
Solamente uno de los ECA examinó lo que se deseaba conocer: la proporción de participantes que tuvieron éxito quirúrgico. No hubo información suficiente disponible para determinar las diferencias entre ninguno de los procedimientos quirúrgicos con respecto al éxito quirúrgico u otro resultado relevante para la revisión. Los eventos adversos más frecuentes de los procedimientos quirúrgicos fueron la desviación hacia abajo del ojo después de la cirugía (hipotropía), el movimiento limitado del ojo hacia arriba y la necesidad de cirugía repetida.

Calidad de la evidencia
Los cuatro estudios incluidos tuvieron deficiencias en el diseño, la realización, o ambos, lo que debilita las conclusiones. Se necesitan ECA bien diseñados y rigurosamente realizados de los tratamientos para la DVD que aporten pruebas más fiables para el tratamiento de esta afección.

Notas de traducción

La traducción y edición de las revisiones Cochrane han sido realizadas bajo la responsabilidad del Centro Cochrane Iberoamericano, gracias a la suscripción efectuada por el Ministerio de Sanidad, Servicios Sociales e Igualdad del Gobierno español. Si detecta algún problema con la traducción, por favor, contacte con Infoglobal Suport, cochrane@infoglobal-suport.com.

Background

Description of the condition

Dissociated vertical deviation (DVD) is an upward drifting of one eye when the other eye is fixing on a target (Brodsky 1999). The deviation often involves both eyes but is most frequently asymmetric, such that the primary concern is vertical drifting of one eye (Helveston 1980). The deviation may be manifest (spontaneously visible to others) or latent (only seen when the eye is covered), and it is the manifest form that leads parents and patients to seek treatment. People with DVD rarely complain of double vision, because the misalignment is associated with central suppression of the image from the deviated eye (the image is not perceived), and parents and patients seek treatment for psychosocial reasons or for subjective strain.

DVD was first described in 1895 by Stevens who called it "anaphoria" or "anatropia" (Wolff 1978). DVD has also been referred to as "alternating hyperphoria," "double hyperphoria," "occlusion hyperphoria," "occlusion hypertropia," "alternating sursumduction," "double dissociated hyperphoria," "dissociated hyperphoria," "dissociated hypertropia," and "dissociated vertical divergence" (Wolff 1978).

DVD most often occurs in the context of pre-existent infantile strabismus, typically infantile esotropia and typically following surgery for esotropia (Arslan 2010). It may occur in up to 90% of cases of infantile esotropia (Helveston 1980; Neely 2001), but also may occur in the context of presumed infantile microstrabismus and exotropia (Choi 2001; Lim 2008). DVD may also occur in association with acquired loss of vision in childhood (Kutluk 1995). The common feature of all these scenarios associated with DVD is "intense disturbance of binocular vision" early in childhood (Houtman 1991).

Not only does DVD present as a visually noticeable vertical ocular misalignment, but it may also result in a noticeable head tilt (Bechtel 1996; Brodsky 2004; Santiago 1998). There is controversy regarding whether DVD can cause symptoms of strain and whether intervention can improve such symptoms.

Although some authors suggest that DVD spontaneously resolves by adulthood (Fleming 1980), other authors report persistence of the condition (Sprague 1980), and no spontaneous improvement (Harcourt 1980), such that many parents seek treatment for their children with DVD, and many adults with DVD also seek treatment. There are no studies evaluating the natural history of DVD.

Some authors have commented that for the person with asymmetric DVD, surgery on only one eye commonly leads to a need for surgery on the other eye, suggesting that asymmetric DVD should be addressed bilaterally (Noel 1982; Sargent 1976).

DVD is one of the least understood types of strabismus. Some investigators believe it is due to a vestigial righting reflex (Brodsky 1999; Brodsky 2002; Brodsky 2011), whereas others feel it is a nystagmus blocking mechanism (Guyton 1998; Guyton 2000; Guyton 2004). Other authors have suggested that DVD is the result of unbalanced cortical input to subcortical pathways (ten Tusscher 2010). Despite lack of clarity regarding the pathophysiology of DVD, many patients undergo empirical treatment.

Description of the intervention

People with DVD are managed with either observation, optical blur to change fixation (Yue 2003), injection of botulinum toxin into the superior rectus (McNeer 1989), or surgery. Reported surgical techniques used for DVD are as follows:

  1. weakening of the inferior oblique (Pratt-Johnson 1976; Strominger 2009)

  2. anteriorization of the inferior oblique muscle (Bacal 1992; Black 1997; Bothun 2004; Burke 1993; Elliott 1981; Engman 2001; Fard 2010; Guemes 1998; Milot 1994; Mims 1999; Nabie 2007; Nelson 2007; Quinn 2000; Stager 1992)

  3. anteriorization of the inferior oblique combined with resection (Farvardin 2002; Quinn 2000; Snir 1999; Wong 2003)

  4. recession of the superior rectus muscle (Braverman 1977; Broniarczyk-Loba 2007; Magoon 1982; McCall 1991; Repka 1988; Schwartz 1991; Scott 1982; Varn 1997; Yu 1992)

  5. recession of the superior rectus muscle with a posterior fixation suture (Esswein 1992; Kii 1994; Lorenz 1992; Sprague 1980)

  6. superior rectus posterior fixation suture alone (Lorenz 1992)

  7. resection of the inferior rectus (Esswein 1994; Noel 1982; Sargent 1976; Sargent 1979)

  8. tucking of the inferior rectus (Arroyo-Yllanes 2007)

  9. four-muscle oblique surgery (Gamio 2002)

  10. superior oblique resection (Richard 1987)

It is noteworthy that a wide spectrum of surgical approaches have been described for DVD, which may indicate a lack of "best practice" agreement among surgeons.

DVD may present with other types of strabismus, for example esotropia and exotropia. This review planned to include studies of participants who also have horizontal strabismus, but exclude participants undergoing interventions concurrent to address horizontal strabismus, or pattern strabismus such as coexistent superior oblique overaction (McCall 1991), since it would not be possible to separate the effects of surgery for horizontal or oblique muscle dysfunction from the effects of surgery for DVD.

How the intervention might work

Non-surgical interventions, such as optical blur of the eye with the less pronounced DVD, may work by promoting fixation of the eye with the more pronounced DVD, thereby making the DVD less noticeable. Strabismus surgery may reduce the tendency for the eye to drift upwards, either by mechanically restricting elevation or by reducing the upward force on the eye.

Why it is important to do this review

Although many people undergo surgical or non-surgical treatment for DVD, whether or not to perform treatment and which type of treatment to administer appear to be based more on the personal preference of the care provider than on good evidence (Caputo 1999; Coats 2011). A comprehensive review is therefore needed to evaluate whether any intervention is effective for DVD and which type of intervention is most effective.

Objectives

The objective of this review was to determine the effectiveness and safety of various surgical and non-surgical interventions in randomized controlled trials (RCTs) of participants with DVD.

Methods

Criteria for considering studies for this review

Types of studies

We included only RCTs in this review. However, as we anticipated that there would be very few RCTs, we ran the searches without a filter in order to identify all reports of studies on this topic (Holmes 2013).

Types of participants

We included trials of participants of any age with DVD undergoing any surgical or non-surgical intervention. We made no exclusions for previous treatment.

Types of interventions

We included any trial comparing surgical versus no intervention or non-surgical intervention, or one surgical intervention compared with another surgical intervention. A set of comparisons include any type of:

  1. surgical intervention versus no intervention;

  2. surgical intervention versus non-surgical intervention;

  3. surgical intervention versus another type of surgical intervention;

  4. non-surgical intervention versus no intervention;

  5. non-surgical intervention versus another type of non-surgical intervention.

Types of outcome measures

Primary outcomes

The primary outcome we specified for comparison of interventions was the proportion of participants with success, defined as reduction of the hyperdeviation in each eye to 4 prism diopters (pd) or less, based on a prism under cover test and assessed at one year from intervention. In the absence of rigorous test retest data for vertical deviations, we elected to use a value of 4 pd or less for success. The total deviation measured in this way was the sum of the DVD and any true hypertropia; it is not currently possible to separate these components (Tarczy-Hornoch 2008). Assessment by prism and alternate cover may yield the same data as the prism under cover test, as long as care is taken to dissociate the participant; therefore we also considered prism and alternative cover test data when prism under cover test data were not available. Wherever prism and alternate cover test data are used, they must be measured separately fixing right eye and fixing left eye.

Secondary outcomes

The secondary outcomes we specified for this review were:

  1. Total magnitude of hyperdeviation, measured with a prism under cover test (or prism and alternative cover test, when prism under cover test data are not available) of each eye separately, summed between both eyes. When prism and alternate cover test are to be used, they must explicitly be collected fixing right eye and then fixing left eye.

  2. Proportion of day (duration) of misalignment (as data are available).

  3. Quality-of-life scores from vision-specific quality-of-life instruments, such as the National Eye Institute (NEI) Visual Function Questionnaire-25 (VFQ-25) for adults, Adult Strabismus-20 Questionnaire (AS-20), and vision-specific pediatric quality-of-life instruments for children.

  4. Proportions of any adverse events that occurred during or after any surgical/non-surgical interventions.

  5. Magnitude of simultaneous prism and cover test (SPCT) of the deviating eye or, if deviating eye not noted, the eye with greatest DVD-type deviation (SPCT at either distance or near fixation)

  6. Magnitude of prism under cover test or prism and alternate cover test of the deviating eye or, if deviating eye not noted, the eye with greatest DVD-type deviation (at either distance or near fixation).

For all outcomes, we planned to analyze data at one year postintervention when possible, but in the absence of one-year data we reported any available data.

Search methods for identification of studies

Electronic searches

We searched CENTRAL (which contains the Cochrane Eyes and Vision Trials Register) (2015, Issue 8), Ovid MEDLINE, Ovid MEDLINE In-Process and Other Non-Indexed Citations, Ovid MEDLINE Daily, Ovid OLDMEDLINE (January 1946 to August 2015), EMBASE (January 1980 to August 2015), PubMed (1948 to August 2015), Latin American and Caribbean Health Sciences Literature Database (LILACS) (1982 to August 2015), the metaRegister of Controlled Trials (mRCT) (www.controlled-trials.com) (last searched 3 Feburary 2014), ClinicalTrials.gov (www.clinicaltrials.gov), and the WHO International Clinical Trials Registry Platform (ICTRP) (www.who.int/ictrp/search/en). We did not use any date or language restrictions in the electronic searches for trials. We last searched the electronic databases on 3 August 2015, with the exception of mRCT which was last searched 3 Feburary 2014.

See: Appendices for details of search strategies for CENTRAL (Appendix 1), MEDLINE (Appendix 2), EMBASE (Appendix 3), PubMed (Appendix 4), LILACS (Appendix 5), mRCT (Appendix 6), ClinicalTrials.gov (Appendix 7), and the ICTRP (Appendix 8).

Searching other resources

We searched the reference lists of relevant studies to identify any further reports of trials that might be eligible for inclusion in the review.

Data collection and analysis

Selection of studies

Two review authors independently screened the titles and abstracts identified from the electronic searches and classified them as "definitely relevant," "possibly relevant," or "definitely not relevant." We retrieved the full-text reports for all records classified as "definitely relevant" or "possibly relevant" by at least one review author. Two review authors independently assessed the eligibility of each study report based on the inclusion criteria of the review. In case of disagreement between review authors, we consulted a third review author in order to reach final agreement. For studies excluded during full-text screening, we recorded the reason for exclusion in the Characteristics of excluded studies table. We contacted or attempted to contact the study investigators of Duncan 1984, Uncovska 2003, and Vodickova 2008 for missing or unclear data, but we did not receive additional information.

Data extraction and management

Two review authors independently extracted data for primary and secondary outcomes using a data collection form developed in collaboration with Cochrane Eyes and Vision. Two review authors recorded study characteristics about methods, participants, and interventions of each included study. After reaching consensus regarding extracted data, one review author entered the data into RevMan 5.3 (RevMan 2014), and a second review author verified the entered data.

Assessment of risk of bias in included studies

Two review authors independently assessed the risk of bias of each included study using the methods described in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011), as part of the data extraction process. We assessed the following domains: selection bias (sequence generation and allocation concealment), performance bias (masking of participants and personnel), detection bias (masking of outcome assessors), attrition bias (incomplete outcome data), reporting bias (selective outcome reporting), and any other potential sources of bias. In case of disagreement in bias assessment, we consulted a third review author. For missing or unclear information, we contacted investigators of Duncan 1984, Uncovska 2003, and Vodickova 2008, but we did not receive additional information, so we assessed the available information. We labeled each domain for each trial as high risk of bias, low risk of bias, or unclear risk of bias.

Measures of treatment effect

The primary outcome for this review, the proportion of participants with success, was a dichotomous outcome. We planned to report the measure of effect as a risk ratio with a 95% confidence interval (CI). We planned to report the dichotomous secondary outcome, proportions of adverse events, in the same manner. For continuous outcomes (all other secondary outcomes), we planned to estimate the mean differences between groups with corresponding 95% CIs when sufficient data were available. We also planned to calculate standardized mean differences when continuous outcomes were measured using different scales.

Unit of analysis issues

The unit of analysis was the eye. Some participants from each trial contributed both eyes to the analyses; however, no consideration was made by the trial authors to account for the non-independence of eyes. For one trial (Quinn 2000), sufficient data were available to reanalyze the data by participant. For the remaining studies, we used the data as available, documented the number of participants contributing both eyes to the data analysis, and discussed the limitations of these data.

In our protocol (Holmes 2013), we planned for the unit of analysis to be the participant for outcomes related to surgical success, total magnitude of hyperdeviation, proportion of day misaligned, and quality-of-life scores, because interventions for DVD may appear to improve the DVD in one eye but worsen the DVD in the opposite eye. We also planned to use individual eye data to classify each eye and then combine these data for a participant-level classification of outcome, because otherwise an intervention that improves the hyperdeviation in one eye but makes the hyperdeviation worse in the fellow eye would erroneously be considered successful.

Dealing with missing data

We contacted investigators of the Duncan 1984, Uncovska 2003, and Vodickova 2008 studies for missing or incomplete data. When the investigators did not respond or did not provide additional information, we used the data that were available. We did not impute data for analysis.

Assessment of heterogeneity

We assessed clinical and methodological heterogeneity by examining potential variations in participant characteristics, inclusion and exclusion criteria, and measurements of primary and secondary outcomes. We also planned to use the I2 statistic (%) to determine the proportion of variation due to statistical heterogeneity, with a value above 50% suggesting substantial statistical heterogeneity.

Assessment of reporting biases

We planned to use funnel plots to assess for publication bias when 10 or more studies were included in a meta-analysis. We assessed selective outcome reporting as part of the 'Risk of bias' assessment.

Data synthesis

We determined whether data synthesis was appropriate depending on the assessment of clinical and methodological heterogeneity. Additionally, if the I2 statistic suggested substantial statistical heterogeneity (above 50%), we planned not to synthesize the data; instead, we would present results in a narrative summary. When the I2 statistic was less than 50% (indicating no substantial statistical heterogeneity), we planned to combine study results using a random-effects model if there were three or more studies. We used a fixed-effect model when the number of studies included in a meta-analysis was fewer than three and there was no evidence of statistical, clinical, or methodological heterogeneity.

Sensitivity analysis

We planned to conduct sensitivity analyses to determine the impact of excluding studies assessed as being at high risk of bias, studies funded by industry, and studies not published in a journal.

Summary of findings

Two review authors independently graded the overall certainty of the evidence for each outcome using the GRADE classification (www.gradeworkinggroup.org/).

Results

Description of studies

See: Characteristics of included studies; Characteristics of excluded studies.

Results of the search

Figure 1 is a study flow diagram documenting the number of references screened and included at each step of the trial selection process. The electronic searches as of 3 August 2015 yielded 2067 titles and abstracts and two records from trial registers. After the Trials Search Co-ordinator removed 871 duplicate records, we screened the remaining 1196 titles and abstracts and two trial register records. After adjudication and exclusion of 1187 total records, we recorded nine records as "definitely relevant" or "possibly relevant" and retrieved their corresponding full-text reports. We reviewed the nine reports from nine studies, and after adjudication, found four studies that met the inclusion criteria of this review (Duncan 1984; Quinn 2000; Uncovska 2003; Vodickova 2008). One study, Vodickova 2008, was translated from Czech into English, and is a potential duplicate of the Uncovska 2003 study with longer follow-up, though we were unable to confirm with the trials authors whether the participants were the same.

Figure 1.

Study flow diagram.

Included studies

We described the study characteristics in detail in the Characteristics of included studies table. None of the studies were registered as clinical trials. Below is a salient summary of all studies.

Types of participants

The four studies included a total of 151 participants (248 eyes), between the age of six months and 22 years and including both genders. One study was conducted in Canada, one in the USA, and two in the Czech Republic. All studies included participants diagnosed with DVD.

Duncan 1984 included both adults and children (ages 16 months to 22 years) who had a minimum magnitude of preoperative DVD of 10 pd (although it is unclear whether or not this was part of the inclusion criteria). It reported three intervention groups, only two of which were prospectively randomized and included in this review; data from the third group were "collected retrospectively" and not eligible for inclusion in this review. The trial authors only reported overall characteristics of all included participants. Individual group data were not available, and baseline characteristics were not compared between intervention groups. Therefore data on participant characteristics from this trial represents all participants reported in the paper. The combined superior rectus recession with posterior fixation group consisted of 19 participants (25 eyes), and the superior rectus recession alone group consisted of 15 participant (17 eyes). Some participants were reported to have had mild to moderate inferior oblique overaction, but it is unclear how many were in each treatment arm.

Quinn 2000 included only children (ages 19 months to 15 years) and specified that DVD measure at least 5 pd in at least one eye for inclusion and require surgery (all operated eyes measured at least 6 pd of DVD). The two intervention groups had similar baseline characteristics (age, gender, alignment complications, and reoperation rates). The combined anteriorization with resection group consisted of 15 participants (25 eyes), and the standard anteriorization group consisted of 15 participants (26 eyes). A proportion of participants in each group (13 (87%) in the combined anteriorization with resection group and 9 (60%) in the standard anteriorization group) had coexisting preoperative inferior oblique muscle overaction, graded from +0.5 to +3.5.

Uncovska 2003 included only children (ages 13 months to 15 years) and required that included participants have 5 pd or more of latent or manifest DVD requiring surgery (all operated eyes measured at least 6 pd of DVD). The anteriorization group consisted of 22 participants (35 eyes), and the myectomy group consisted of 19 participants (38 eyes). Whether or not baseline characteristics were similar between the two groups was unclear. All children had previous or planned horizontal strabismus surgery.

Vodickova 2008 included only children (ages 13 months to 15 years), possibly the same children as Uncovska 2003, and required manifest or latent DVD of 5 pd or more for inclusion. The anteriorization group consisted of 27 participants (44 eyes), and the myectomy group consisted of 19 participants (38 eyes). Whether or not the baseline characteristics were similar between the two groups was unclear.

Types of interventions

We did not identify any studies of non-surgical interventions. The four studies compared different types of surgical interventions. Duncan 1984 compared superior rectus recession of 3 to 5 mm with versus without a posterior fixation suture at 12 to 14 mm. Participants were followed for an average of 19 months (range 5 to 84 months). Quinn 2000 compared anteriorization of the inferior oblique muscle with versus without a 7 mm resection. Participants were followed for an average of 15.5 months (range 6 to 39 months) and 25.2 months (range 4 to 42 months) in the 2 groups, respectively. Uncovska 2003 compared anterior transposition of the inferior oblique with myectomy of the inferior oblique and followed participants for an average of 19.6 months (range 7 to 43 months) and 26.5 months (range 8 to 52 months), respectively. Vodickova 2008 also compared anterior transposition of the inferior oblique with myectomy of the inferior oblique (possibly the same participants as Uncovska 2003). Participants were followed for an average of 29.6 months (range 7 to 53 months) and 36.5 months (range 8 to 72 months), respectively.

None of the four studies reported specific follow-up time points for the reporting of primary and secondary outcomes. Presumably, for each study, data reported were taken from participants' most recent follow-up visits.

Types of outcomes

The included studies reported different types of outcomes. Duncan 1984 reported the proportion of participants in each group classified as cured, improved, or unchanged. "Cure" was defined as "latency or elimination of hyperdeviation," and "improved" as "intermittent but seldom noted, or remained manifest but decreased in size to a cosmetically acceptable angle." Neither of these definitions included quantification of magnitude or frequency. Quinn 2000 reported the magnitude of postoperative DVD in prism diopters in each treatment group as well as the grade of inferior oblique overaction and superior oblique overaction, and the presence or absence of a true hypotropia in the primary position. Uncovska 2003 compared mean magnitude of change in pre- to postoperative DVD in pd within each treatment group, but not between treatment groups. For eyes with more than 15 pd of preoperative DVD, postoperative angle of DVD was used to classify outcomes as either excellent (0 to 4 pd), good (5 to 9 pd), fair (10 to 14 pd), or poor (more than 14 pd). The proportions of eyes in each of these four categories were compared between treatment groups. Nevertheless, data were only reported for the operated eye, and not for the unoperated eye, thus not meeting the criteria for the primary outcome of this review (and highlighting missing data). We included operated eye data under secondary outcomes for this trial. In addition, the grade of pre- to postoperative inferior oblique overaction was compared within groups, but not between groups. Vodickova 2008 compared pre- to postoperative change in median magnitude of DVD in pd within each treatment group, across all participants. There was no overall comparison of the treatment groups to each other. As in the Uncovska 2003 study, eyes with more than 15 pd of DVD preoperatively were classified as either excellent (0 to 4 pd), good (5 to 9 pd), fair (10 to 14 pd), or poor (more than 14 pd) postoperatively. Again, data were reported only for the operated eye and not for the unoperated eye, thus not meeting the criteria for the primary outcome of this review (but were included under secondary outcomes). The pre- to postoperative grade of inferior oblique overaction was compared within, but not between, treatment groups.

All studies reported adverse events. None of the included studies reported data for the following secondary outcomes specified for our review: proportion of day (duration) of misalignment, quality-of-life scores, magnitude of simultaneous prism or cover test of the deviating eye, and magnitude of prism under cover test or prism and alternate cover test.

Funding sources

Duncan 1984, Uncovska 2003, and Vodickova 2008 did not report funding sources. Quinn 2000 was supported in part by a grant from the Glaucoma Trust of the Ophthalmological Society of New Zealand.

Excluded studies

We excluded five studies and listed the reasons for exclusion in the Characteristics of excluded studies table. Four studies were not RCTs (Apers 1981; Castellanos-Bracamontes 1997; Kratz 1989; Velez 1993), and one had unclear participant eligibility criteria (Escalante-Razo 2002).

Risk of bias in included studies

See: Figure 2

Figure 2.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study. Yellow circles indicate the risk of bias is unclear; red circles indicate a high risk of bias.

Allocation

None of the studies reported the scheme they used for random sequence generation or allocation concealment. We judged all studies to have unclear risk of selection bias.

Masking (performance bias and detection bias)

None of the studies reported whether they masked participants. Due to the nature of surgical trials, surgeons cannot be masked. We judged all studies to have unclear risk of performance bias.

Duncan 1984 did not report how outcomes were assessed, and so we judged this study to have unclear risk of detection bias. Quinn 2000 reported that when possible, the outcome assessment was performed by an orthoptist who was masked to the procedure performed. This occurred in 13 of 15 participants (87%) in the combined anteriorization with resection group and in 4 of 15 participants (27%) in the standard anteriorization group. The reason for a greater proportion of masked exams in the combined anteriorization with resection group compared with the standard anteriorization group is unclear. Given the disproportionate number of masked exams in the two groups, we judged there to be a high risk of detection bias in this study. Uncovska 2003 also reported that "where possible" postoperative measurements were taken by an orthoptist who was unaware of the surgical procedure. The authors did not report how many participants had a masked outcome assessment, and we therefore judged the study to have an unclear risk of detection bias. Vodickova 2008 did not report how outcomes were assessed, so we judged this study to have an unclear risk of detection bias.

Incomplete outcome data

None of the studies reported loss to follow-up at the last follow-up visit because none of the studies had a standardized outcome time point. Since the outcome time points were not specified, we judged each of these studies to have an unclear risk of attrition bias. All the included studies reported outcomes at only one time point, the last follow-up, but participants may well have been assessed more frequently. Duncan 1984 reported overall follow-up as ranging from 5 to 84 months for all three study groups combined, whereas Quinn 2000 reported that all participants were followed for a minimum of 4 months. Uncovska 2003 reported follow-up separately for each group (ranging from 7 to 52 months). Vodickova 2008 also reported follow-up separately for each group with an overall range from 7 to 72 months.

Selective reporting

We judged all studies to have an unclear risk of reporting bias due to insufficient information.

Other potential sources of bias

For the Duncan 1984 study, we did not identify other potential sources of bias, but our ability to do so was limited by inadequate or unclear reporting. In the Quinn 2000 study, the average follow-up was much longer in the resection group (mean 25.2 months) than in the standard anteriorization group (mean 15.5 months), potentially confounding comparison of outcomes, so we also judged this to be unclear risk of bias.

In the Uncovska 2003 study, there was a discrepancy in reported postoperative measurements. Across all included participants the postoperative DVD ranged from 0 to 14 pd in the anteriorization group and from 0 to 13 pd in the myectomy group. However, in the reporting of postoperative measurements on the subgroup of eyes with more than 15 pd preoperatively, there were two eyes (one in each treatment group) with more than 14 pd of DVD postoperatively. This discrepancy suggests that either the reported ranges for the whole cohort are incorrect, or there was an error in the reporting of outcome data in eyes with more than 15 pd DVD preoperatively. Such discrepancies undermine the study conclusions.

Vodickova 2008 is suspect for not being an independent RCT since the number of participants, baseline characteristics, and postoperative alignment for the myectomy group are exactly the same as those reported in the Uncovska 2003 study (both studies were conducted in the Czech Republic). The only differences between the Uncovska 2003 and Vodickova 2008 studies regarding included participants was that the Vodickova 2008 study had an additional five participants in the anteriorization group, and the Vodickova 2008 study reported slightly longer follow-up. The Vodickova 2008 study authors made no reference to the Uncovska 2003 study; the Vodickova 2008 study was reported as an original RCT. We consider it highly unlikely that an independent RCT would include participants with exactly the same demographic characteristics and outcomes in one of the treatment groups as those in a previous RCT. In addition, in both studies there were inconsistencies regarding the number of eyes in each group, and in the Vodickova 2008 study, the number of participants in mutually exclusive subgroups did not add up to the stated total, raising further concerns regarding the reliability of the reported data. However, without clarification from the authors and lacking more definitive evidence for the Vodickova 2008 study being a duplicate publication, we still included it in this review.

Effects of interventions

Among the four studies included, there were three different surgical intervention comparisons. The Duncan 1984 study compared superior rectus recession with posterior fixation suture versus superior rectus recession alone. The Quinn 2000 study compared anteriorization of the inferior oblique muscle with 7 mm resection versus anteriorization of inferior oblique muscle alone. Two studies, Uncovska 2003 and Vodickova 2008, compared anteriorization of the inferior oblique muscle versus myectomy of the inferior oblique muscle. None of the studies reported outcomes at a specified follow-up time, rather outcomes were reported using the last follow-up measurement (range 4 to 84 months). The unit of analysis for all studies was the eye; however, Quinn 2000 reported individual participant data, for which we were able to analyze outcomes by participant.

Superior rectus recession with posterior fixation suture versus superior rectus recession alone

Proportion of eyes with success

Although Duncan 1984 reported proportions of eyes "cured," defined as "latency or elimination of hyperdeviation," it was unclear how many had "elimination of the hyperdeviation" and how many had "latency" of the deviation. Since we could not be certain that any latent deviation measured 4 pd or less, we were unable to determine how many eyes met our primary outcome. There were 14 eyes (56%) considered "cured" in the superior rectus recession with posterior fixation suture group and 6 eyes (35%) considered "cured" in the superior rectus recession alone group; the difference in eyes that were "cured" between the two groups was uncertain (risk ratio (RR) 1.59, 95% confidence interval (CI) 0.76 to 3.30, graded very low quality due to study limitations and imprecision).

Total magnitude of hyperdeviation

Duncan 1984 did not report total magnitude of hyperdeviation.

Proportion of day (duration) of misalignment

Duncan 1984 did not report the proportion of day (duration) of misalignment.

Quality-of-life scores

Duncan 1984 did not report quality-of-life outcomes.

Adverse events

Four eyes had postoperative hypotropia in the primary position, two (8%) in the combined superior rectus recession with posterior fixation group, and two (12%) in the superior rectus recession alone group (RR 0.68, 95% CI 0.11 to 4.37, graded very low quality).

No eyes in the combined superior rectus recession with posterior fixation group had limitation of elevation compared with two (12%) in the superior rectus recession alone group (RR 0.14, 95% CI 0.01 to 2.72, graded very low quality). One of the eyes with limitation of elevation had undergone prior inferior rectus resection.

Magnitude of simultaneous prism and cover test (SPCT) of the deviating eye

Duncan 1984 did not report SPCT outcomes.

Magnitude of prism under cover test or prism and alternate cover test of the deviating eye

Duncan 1984 did not report prism under cover test or prism and alternate cover test outcomes.

Anteriorization of the inferior oblique muscle with 7 mm resection versus anteriorization of the inferior oblique muscle alone

Proportion of eyes with success

Quinn 2000 reported the postoperative angle of deviation for each eye, albeit measured using the prism and alternate cover test rather than the prism under cover test, and so we could classify data according to the primary outcome of success for this review. Quinn 2000 stated that outcomes were to be measured at least four months postoperatively, but data were analyzed for the last available follow-up. Therefore the actual outcome time point varied from participant to participant: in the anteriorization with resection group it ranged from 4 to 42 months (average 25.2), and in the standard anteriorization group it ranged from 6 to 39 months (average 15.5). In the group that received anteriorization with resection, 9 of 15 participants (60%) achieved success, while in the standard anteriorization group the surgery was successful in 8 of 15 participants (53%); there was uncertainty in the effect of interventions between groups (RR 1.13, 95% CI 0.60 to 2.11, graded very low quality due to study limitations and imprecision).

Total magnitude of hyperdeviation

Total magnitude of hyperdeviation (each eye summed) was measured using the prism and alternate cover test. The mean magnitude of hyperdeviation in the combined anteriorization with resection group was 7.6 pd (median 7 pd, range 0 to 24 pd), and the mean magnitude of hyperdeviation for the standard anteriorization group was 9 pd (median 8 pd, range 0 to 20 pd).

Proportion of day (duration) of misalignment

Quinn 2000 did not report the proportion of day (duration) of misalignment.

Quality-of-life scores

Quinn 2000 did not report quality-of-life outcomes.

Adverse events

The study reported that three (12%) participants in the combined anteriorization with resection group versus two (8%) participants in the standard anteriorization group developed primary position hypotropia postoperatively (RR 1.50, 95% CI 0.29 to 7.73, graded very low quality). One additional participant in the standard anteriorization group had a primary position hypotropia preoperatively.

In addition, six eyes of three participants (20%) in the combined anteriorization with resection group developed mild (-0.5 to -1) limitation of elevation in abduction postoperatively, whereas none of the participants in the standard anteriorization group developed this complication (RR 7.00, 95% CI 0.39 to 124.83, graded low quality).

Magnitude of SPCT of the deviating eye

Quinn 2000 did not report SPCT outcomes.

Magnitude of prism under cover test or prism and alternate cover test of the deviating eye

Quinn 2000 did not report prism under cover test or prism and alternate cover test outcomes by deviating eye or by eye with the largest deviation.

Anteriorization of the inferior oblique muscle versus myectomy of the inferior oblique muscle

Proportion of eyes with success

Two studies reported data on inferior oblique anteriorization versus inferior oblique myectomy, but only recorded the postoperative deviation in the operated eye, precluding inclusion of data for the primary analysis: Uncovska 2003 included 41 participants (73 eyes), and Vodickova 2008 (which may in fact be the same study population as Uncovska 2003) included 46 participants (82 eyes).

Total magnitude of hyperdeviation

Both Uncovska 2003 and Vodickova 2008 reported postoperative angle of deviation in the operated eye only. For each eye, data were collected at one postoperative time point, which for the anteriorization groups ranged from 7 to 43 (average 19.6) months postoperatively in the Uncovska 2003 study and from 7 to 53 (average 29.6) months postoperatively in the Vodickova 2008 study. For myectomy groups, data were collected at follow-up time points that ranged from 8 to 52 (average 26.5) months postoperatively in the Uncovska 2003 study and from 8 to 72 (average 36.5) months postoperatively in the Vodickova 2008 study.

Across all eyes, data were reported only as mean or median (range) angle of deviation in pd. Among eyes with preoperative DVD greater than 15 pd, results were classified as either excellent (0 to 4 pd), good (5 to 9 pd), fair (10 to 14 pd), or poor (greater than 14 pd). We combined data from both studies assuming they were independent studies to compare excellent outcomes for anteriorization versus myectomy (RR 0.93, 95% CI 0.63 to 1.37, graded very low quality due to study limitations and imprecision; Analysis 1.1).

Proportion of day (duration) of misalignment

Uncovska 2003 and Vodickova 2008 did not report the proportion of day (duration) of misalignment.

Quality-of-life scores

Uncovska 2003 and Vodickova 2008 did not report quality-of-life outcomes.

Adverse events

Both studies reported the number of eyes that had limited elevation in abduction following surgery and the number of eyes needing repeat surgery. When data were combined, fewer eyes in the anteriorization group experienced limited elevation in abduction following surgery compared with eyes in the myectomy group (RR 0.19, 95% CI 0.04 to 0.85; Analysis 1.2); however, we graded these results as low quality because the data may represent the same study population. The difference in the need for repeat surgery between anteriorization and myectomy groups was uncertain (RR 0.58, 95% CI 0.22 to 1.52; graded very low quality; Analysis 1.2).

Magnitude of SPCT of the deviating eye

Uncovska 2003 and Vodickova 2008 did not report SPCT outcomes.

Magnitude of prism under cover test or prism and alternate cover test of the deviating eye

Uncovska 2003 and Vodickova 2008 did not report prism under cover test or prism and alternate cover test outcomes by deviating eye or by eye with the largest deviation.

Discussion

Dissociated vertical deviation (DVD) is a unique eye movement disorder for which there are a number of different surgical and non-surgical treatment options. Although there are a number of studies comparing different treatment approaches, the majority are not randomized trials. We found four randomized trials that were eligible for inclusion in this review. We suspect that two of the studies may be potential duplicates, but do not have data to independently confirm one way or the other, which makes it difficult to draw definitive conclusions about the surgical treatment options evaluated in these two trials (Uncovska 2003; Vodickova 2008).

Summary of main results

The four included trials evaluated five different surgical approaches for the correction of DVD: inferior oblique anteriorization, inferior oblique anteriorization plus resection, superior rectus recession, superior rectus recession plus posterior fixation suture, and inferior oblique myectomy. Only one study provided data that allowed analysis of the primary outcome of this review (Quinn 2000). Overall, the limited outcome data preclude drawing definitive conclusions regarding optimal surgical treatment.

One study evaluated the effectiveness of superior rectus recession with posterior fixation suture versus superior rectus recession (Duncan 1984), but the reported data did not allow evaluation of our review outcomes. This study reported the number of participants who were "cured," but they did not differentiate between "elimination of the hyperdeviation" and "latency" of the hyperdeviation, or quantify these descriptions. Since our primary outcome of interest was reduction of hyperdeviation to 4 pd or less, we could not include these data.

Quinn 2000 reported that inferior oblique anteriorization combined with a 7 mm resection and standard inferior oblique anteriorization were similarly effective in correcting DVD with respect to surgical success and the total (summed) hyperdeviation after surgery, although there was uncertainty in the results due to a small number of participants and variation in follow-up times.

Two additional studies evaluated the effect of inferior oblique anteriorization versus inferior oblique myectomy for the correction of DVD, although we suspect these studies are not independent (Uncovska 2003; Vodickova 2008). However, in both studies the authors did not report the amount of DVD present in the non-operated eye, precluding use of their outcome data for the primary analysis. The amount of DVD in the non-operated eye may have been greater than 4 pd postoperatively. In analyses of the magnitude of hyperdeviation (a secondary outcome), we evaluated data from Uncovska 2003 and Vodickova 2008 for participants with preoperative DVD greater than 15 pd (combined total 61 eyes). A postoperative deviation of 0 to 4 pd in the operated eye was achieved in 18 of 30 eyes (60%) in the anteriorization group compared with 20 of 31 eyes (65%) in the myectomy group.

The most common adverse events reported across surgeries for DVD were development of primary position hypotropia postoperatively, limited elevation in abduction following surgery, and the number of eyes needing repeat surgery.

We had planned to conduct sensitivity analyses; however, as we identified three different comparisons of surgeries for DVD and only one study provided data for our primary outcome, a sensitivity analysis was not feasible or necessary.

Overall completeness and applicability of evidence

The studies included in this review evaluated mostly children, and some adults, with varying degrees of DVD severity and lengths of postoperative follow-up.

For all included studies the length of postoperative follow-up was variable; there was no prespecified time point for outcome assessments for any of the studies. In Duncan 1984, follow-up varied from 5 to 84 months and was not reported separately for each treatment group. In Quinn 2000, the length of postoperative follow-up was comparable between treatment groups, but it varied widely for individual participants within each group: range 4 to 42 months for the resection group and 6 to 39 months for the standard group. Uncovska 2003 and Vodickova 2008 also had highly variable follow-up (range 7 to 43 months and 7 to 53 months in the anteriorization group, respectively; and 8 to 52 months and 8 to 72 months in the myectomy group, respectively). DVD may recur over time, and it is possible that the successful outcomes were in those participants with shorter follow-up. The wide variability in the length of follow-up in both studies reduces our ability to interpret the results and determine their applicability.

Quinn 2000 described well and Uncovska 2003 and Vodickova 2008 described reasonably well inclusion and exclusion criteria and baseline characteristics; these three studies only included children. In Duncan 1984, the study population was poorly described with no information regarding age or severity of disease in each treatment group. In addition, there may be differences in treatment outcomes in children and adults, but results were not segregated by age. The lack of such data makes it difficult to apply the results of the Duncan 1984 study to either child or adult populations with DVD.

All four studies reported angle of deviation as quantified using the standard prism and alternate cover test (not the preferred prism under cover test). Some younger participants (number not reported) in the Uncovska 2003 and Vodickova 2008 studies were measured using the Krimsky test, which is less accurate than the prism and alternate cover test.

Although the included studies span a period of nearly 25 years, there has been little change to surgical technique or background therapies during that time, and we therefore do not consider there to be any impact of these potential influences on study findings. Evidence for the optimal management of DVD is far from complete, and several key questions remain unanswered, highlighting the pressing need for further research.

Quality of the evidence

We graded the overall quality of the evidence as very low to low, primarily because of poor reporting of study methods to adequately assess risk of bias, unclear definition of outcomes, lack of standard time points for outcome assessment, unit of analysis issues (eyes treated as independent units), and uncertainty of results. Only one trial reported data relevant to our primary outcome, the proportion with surgical success (Quinn 2000). For all four studies, the variable length of follow-up is of great concern. In Duncan 1984, the overall range was from 5 to 84 months; in Quinn 2000, the range was from 4 to 42 months; in Uncovska 2003, the range was from 7 to 52 months; and in Vodickova 2008, the range was 7 to 72 months. Since the magnitude of DVD may change over time and DVD may recur after intervention, there may be a bias towards finding success in the participants with the least amount of follow-up and failure in those with more. In addition, the average follow-up in Quinn 2000 was much longer for the resection group than the standard anteriorization group, further confounding comparison of outcomes.

The quality of evidence in Duncan 1984 was very low overall. There was no quantification of the angle of deviation postoperatively, which not only precluded applying the primary outcome criteria for this review, but made it difficult to truly estimate the effectiveness of each of the procedures. In addition, "cure" was defined as the proportion of eyes with "latent" versus "eliminated" hyperdeviation, but these two subgroups of "cured" participants were not separated. Although all participants included in Duncan 1984 showed spontaneous DVD preoperatively, such spontaneous manifestation of DVD is variable. It is possible that participants were classified as "cured" based purely on whether they happened to not show manifest DVD during the postoperative exam, and that the angle of deviation was in fact unchanged or little improved. The lack of information regarding the characteristics of participants in each treatment group allows for the possibility that baseline differences may explain the differences in treatment outcomes. The trial authors presented no statistical analysis of the data, and it is unknown whether the outcome assessment was masked. The participant group was a mixture of adults and children, and treatment outcomes may differ for each of these subgroups. As in all of the included studies, the method of randomization was not described, there was no standardized time point for the outcome assessment, and eyes, rather than individuals, were the unit of analysis.

The quality of evidence in Quinn 2000 was low overall. Inclusion and exclusion criteria were well described, and postoperative outcomes were quantified using a standard clinical test (the prism and alternate cover test), although not the preferred prism under cover test. However, there was no standardized time point for the outcome assessment, and there was a difference in the proportion of masked exams in the 2 groups: 13 of 15 participants (87%) in the resection group and 4 of 15 participants (27%) in the standard anteriorization group underwent a masked exam. Although the unit of analysis was the eye, individual participant data were available so that we were able to analyze outcomes by participant for this review.

In Uncovska 2003, the quality of evidence was very low overall. Although DVD was measured using the prism and alternate cover test pre- and postoperatively, and basic inclusion and exclusion criteria were described, some critical details were not reported (for example there was no information regarding the proportion of masked outcome assessments performed or the method of randomization, and there was no standardized time point for the outcome assessment). In addition, postoperative outcomes of the two randomized groups were not formally compared for eyes with DVD measuring 15 pd or less preoperatively (the majority of the trial cohort), but were only compared for those with preoperative DVD measuring 15 pd or more. There also is a data discrepancy in reported postoperative measurements. One table stated the overall range of postoperative DVD to be from 0 to 14 pd in the anteriorization group and from 0 to 13 pd in the myectomy group. However, another table showed that two eyes, one in each treatment group, had greater than 14 pd of DVD postoperatively.

The quality of evidence in Vodickova 2008 also was very low overall, and we have concerns that this trial was not independent of Uncovska 2003. Similar to Uncovska 2003, the basic inclusion and exclusion criteria were described, but there was no information regarding whether there was a masked assessment of outcome or how participants were randomized. There was no standardized time point for the outcome assessment. In addition, postoperative outcomes of the two randomized groups were not formally compared for eyes with DVD measuring 15 pd or less preoperatively (the majority of the trial cohort), but were only compared for those with preoperative DVD measuring 15 pd or more. There also were discrepancies in the denominator for the numbers of eyes in each group reported in the text and tables of the paper.

Potential biases in the review process

We had to decide whether or not to include Vodickova 2008 as an independent study, when we had a strong suspicion that it was a duplicate of the Uncovska 2003 study. This decision could be subject to bias. Since we were unable to obtain clarification regarding whether or not Vodickova 2008 was a duplicate study, we decided that there was less bias in including the study than there was in excluding it, as we did not want to miss potential information about the comparison of anteriorization of the inferior oblique muscle versus myectomy of the inferior oblique muscle. We are unaware of any other potential biases in the review process.

Agreements and disagreements with other studies or reviews

The existing literature on the management of DVD consists largely of case reports, retrospective case series, or cohort studies where a single intervention was evaluated, or two interventions were compared in a non-randomized fashion. The types of bias present in these studies, along with the range of outcome measures used and the different inclusion and exclusion criteria, all preclude useful comparison with the studies included in this review. We did not identify any previous reviews of the management of DVD.

In current practice, the most frequently performed surgical approaches appear to be superior rectus weakening and inferior oblique anteriorization. Choice of procedure often appears to be driven by the personal preference of the care provider, or the common practices of where the surgeon trained.

Across existing studies, including the randomized studies included in this review, there remains no compelling evidence regarding which type of surgical or non-surgical intervention is most effective for the management of DVD. Although data are limited, they do not suggest notable differences in complication rates between the various surgical procedures.

Authors' conclusions

Implications for practice

There was only one study with primary outcome data for this review (Quinn 2000). While this study ostensibly provides some guidance regarding optimal surgical procedures, there were considerable quality issues and methodological concerns that undermined the study's conclusion that anteriorization of the inferior oblique muscle with or without resection provide comparable outcomes. In addition, the trials investigating the more common approaches of superior rectus recession with or without posterior fixation suture, and inferior oblique anteriorization versus inferior oblique myectomy suffered from the same quality and methodological shortcomings.

There were no studies comparing surgical intervention to no surgical intervention. Although anecdotally and in non-randomized case series, surgical intervention often appears to reduce the degree or frequency of DVD, there is currently no high quality evidence to support or refute this suggestion.

There remains a significant lack of reliable evidence to inform best practice, and a pressing need for studies evaluating the effectiveness of various surgical and non-surgical interventions.

Implications for research

There is a clear need for more reliable evidence on the management of DVD. Future clinical trials should include measurement of DVD in both eyes (regardless of which eye(s) underwent treatment), using a standardized measurement of the deviation such as the prism under cover test, to ensure similar levels of dissociation for each eye, and across participants. A rigorous definition of treatment success should be applied (such as 4 pd or less, as proposed in this review) and trials should aim to evaluate long-term outcomes in light of the possibility of recurrence of DVD, and assess health-related quality of life.

Questions for randomized controlled trials that we consider to be of particular importance are:

  1. How effective are commonly performed surgical procedures for DVD? Need for studies including untreated control groups as well as studies directly comparing treatments.

  2. How effective are non-surgical treatments for DVD? Need for studies including untreated control groups as well as studies directly comparing two treatments.

  3. What is the natural history of DVD?

  4. What are the effects of DVD on health-related quality of life?

Acknowledgements

We acknowledge Lori Rosman, Cochrane Eyes and Vision (CEV) Trials Search Co-ordinator, for developing the search strategy and executing the electronic searches. We also acknowledge the CEV editorial team and the peer reviewers for their support and comments during the preparation of this review.

Data and analyses

Download statistical data

Comparison 1. Anteriorization of the inferior oblique muscle vs myectomy of the inferior oblique muscle
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Excellent results (0 to 4 prism diopters) after surgery261Risk Ratio (M-H, Fixed, 95% CI)0.93 [0.63, 1.37]
2 Adverse events2 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
2.1 Limited elevation2155Risk Ratio (M-H, Fixed, 95% CI)0.19 [0.04, 0.85]
2.2 Repeat surgery2155Risk Ratio (M-H, Fixed, 95% CI)0.58 [0.22, 1.52]
Analysis 1.1.

Comparison 1 Anteriorization of the inferior oblique muscle vs myectomy of the inferior oblique muscle, Outcome 1 Excellent results (0 to 4 prism diopters) after surgery.

Analysis 1.2.

Comparison 1 Anteriorization of the inferior oblique muscle vs myectomy of the inferior oblique muscle, Outcome 2 Adverse events.

Appendices

Appendix 1. CENTRAL search strategy

#1 Dissociated vertical deviation*
#2 Anaphoria*
#3 Anatropia*
#4 Alternating hyperphoria*
#5 Double hyperphoria*
#6 Occlusion hyperphoria*
#7 Occlusion hypertropia*
#8 Alternating sursumduction*
#9 Dissociated hyperphoria*
#10 Dissociated hypertropia*
#11 Vertical deviation*
#12 Vertical divergen*
#13 Dissociated hyperdeviation*
#14 #1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9 or #10 or #11 or #12 or #13

Appendix 2. MEDLINE (Ovid) search strategy

1. Dissociated vertical deviation*.tw.
2. Anaphoria*.tw.
3. Anatropia*.tw.
4. Alternating hyperphoria*.tw.
5. Double hyperphoria*.tw.
6. Occlusion hyperphoria*.tw.
7. Occlusion hypertropia*.tw.
8. Alternating sursumduction*.tw.
9. Dissociated hyperphoria*.tw.
10. Dissociated hypertropia*.tw.
11. Vertical deviation*.tw.
12. Vertical divergen*.tw.
13. Dissociated hyperdeviation*.tw.
14. 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13

Appendix 3. EMBASE.com search strategy

#1 anaphoria*:ab,ti
#2 anatropia*:ab,ti
#3 (alternating NEXT/1 hyperphoria*):ab,ti
#4 (double NEXT/1 hyperphoria*):ab,ti
#5 (occlusion NEXT/1 hyperphoria*):ab,ti
#6 (occlusion NEXT/1 hypertropia*):ab,ti
#7 (alternating NEXT/1 sursumduction*):ab,ti
#8 (dissociated NEXT/1 hyperphoria*):ab,ti
#9 (dissociated NEXT/1 hypertropia*):ab,ti
#10 (vertical NEXT/1 deviation*):ab,ti
#11 (vertical NEAR/1 divergen*):ab,ti
#12 (dissociated NEXT/1 hyperdeviation*):ab,ti
#13 'dissociated vertical deviation':ab,ti OR 'dissociated vertical deviations':ab,ti
#14 #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13

Appendix 4. PubMed search strategy

#1 Dissociated vertical deviation*[tw] NOT Medline[sb]
#2 Anaphoria*[tw] NOT Medline[sb]
#3 Anatropia*[tw] NOT Medline[sb]
#4 Alternating hyperphoria*[tw] NOT Medline[sb]
#5 Double hyperphoria*[tw] NOT Medline[sb]
#6 Occlusion hyperphoria*[tw] NOT Medline[sb]
#7 Occlusion hypertropia*[tw] NOT Medline[sb]
#8 Alternating sursumduction*[tw] NOT Medline[sb]
#9 Dissociated hyperphoria*[tw] NOT Medline[sb]
#10 Dissociated hypertropia*[tw] NOT Medline[sb]
#11 Vertical deviation*[tw] NOT Medline[sb]
#12 Vertical divergen*[tw] NOT Medline[sb]
#13 Dissociated hyperdeviation*[tw] NOT Medline[sb]
#14 #1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9 or #10 or #11 or #12 or #13

Appendix 5. LILACS search strategy

Anaphoria$ OR Anatropia$ OR "Alternating hyperphoria" OR "Alternating hyperphorias" OR "Double hyperphoria" OR "Double hyperphorias" OR "Occlusion hyperphoria" OR "Occlusion hyperphorias" OR "Occlusion hypertropia" OR "Occlusion hypertropias" OR "Alternating sursumduction" OR "Alternating sursumductions" OR "Dissociated hyperphoria" OR "Dissociated hyperphorias" OR "Dissociated hypertropia" OR "Dissociated hypertropias" OR "vertical divergent" OR "vertical divergence" OR "vertical deviation" OR "vertical deviations" OR "Dissociated hyperdeviation" OR "Dissociated hyperdeviations" OR "dissociated vertical deviation" OR "dissociated vertical deviations"

Appendix 6. metaRegister of Controlled Trials search strategy

Dissociated vertical deviation OR Dissociated vertical divergence

Appendix 7. ClinicalTrials.gov search strategy

Dissociated vertical deviation OR Dissociated vertical divergence

Appendix 8. ICTRP search strategy

Dissociated vertical deviation OR Dissociated vertical divergence

What's new

DateEventDescription
11 January 2016AmendedMinor edit made to Effects of interventions section

Contributions of authors

Conceived the review: JMH
Designed the review: JMH, XW
Developed and executed searches: Lori Rosman, Cochrane Eyes and Vision
Screened search results: SRH, XW
Organized retrieval of papers: SRH, XW
Screened retrieved papers against inclusion criteria: SRH, XW
Appraised risk of bias: SRH, XW
Abstracted data from papers: SRH, XW
Wrote to authors of papers for additional information: SRH, XW
Provided additional data about papers: SRH, XW, JMH
Obtained and screened data on unpublished studies: SRH, XW
Managed data for the review: SRH, XW
Entered data into RevMan: SRH, XW
Analyzed data: SRH, XW
Interpreted data: SRH, XW, JMH
Wrote the review: SRH, XW, JMH
Performed previous work that was the foundation of current study: SRH, XW, JMH
Guarantor for the review: SRH

Declarations of interest

SRH: none known.
XW: none known.
JMH: none known.

Sources of support

Internal sources

  • No sources of support supplied

External sources

  • National Eye Institute, National Institutes of Health, USA.

    Jonathan M. Holmes and Sarah R. Hatt are partly funded by National Institutes of Health Grant EY018810 and Grant EY024333

  • National Eye Institute, National Institutes of Health, USA.

    Xue Wang is supported by the Cochrane Eyes and Vision - US Project through the National Eye Institute Grant 1 U01 EY020522

  • National Institute for Health Research (NIHR), UK.

    • Richard Wormald, Co-ordinating Editor for the Cochrane Eyes and Vision (CEV) acknowledges financial support for his CEV research sessions from the Department of Health through the award made by the National Institute for Health Research to Moorfields Eye Hospital National Health Service (NHS) Foundation Trust and University College London Institute of Ophthalmology for a Specialist Biomedical Research Centre for Ophthalmology.

    • The NIHR also funds the CEV Editorial Base in London.

    The views expressed in this publication are those of the authors and not necessarily those of the NIHR, NHS, or the Department of Health.

Differences between protocol and review

In our protocol (Holmes 2013), we planned for the unit of analysis to be the participant; however, because of how trial data were reported, the unit of analysis for this review was the eye. Due to a limited number of included trials and insufficient reporting of data, we did not conduct all data synthesis methods described in the protocol for this review. In accordance with Cochrane standards, we modified our methods to include assessment of the certainty of the evidence using the GRADE classification (www.gradeworkinggroup.org/).

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Duncan 1984

Methods

Study design: parallel-group RCT

Number randomized: 34 total participants (42 eyes); 19 participants (25 eyes) in the combined superior rectus recession with posterior fixation group and 15 participants (17 eyes) in the superior rectus recession alone group

Exclusions after randomization: none reported

Losses to follow-up: unclear; minimum of 5 months follow-up for all participants, but data analyzed at what seems to be the last available follow-up for each participant (i.e. no designated outcome time point at which to assess any loss to follow-up)

Number analyzed: 34 total participants (42 eyes); 19 participants (25 eyes) in the combined superior rectus recession with posterior fixation group and 15 participants (17 eyes) in the superior rectus recession alone group

Unit of analysis: the eye (8 participants had both eyes analyzed)

How was missing data handled?: no missing data reported for primary and secondary outcomes

Power calculation: none reported

Other study design details: the randomized groups also were compared with a retrospective cohort (not included in this review). The included cohorts were prospectively recruited and randomized. 4 participants were included in 2 different surgical groups, and the second procedure was performed after the minimum follow-up time and each result was evaluated independently

Participants

Country: USA

Age: mean 8 years (range 16 months to 22 years) including randomized participants and non-randomized participants not included in this review; not reported by groups

Gender: not reported

Inclusion criteria: DVD with or without inferior oblique overaction

Exclusion criteria: none reported

Equivalence of baseline characteristics: not reported

Interventions

Intervention 1: combined superior rectus recession (3 to 5 mm) with posterior fixation suture (12 to 14 mm)

Intervention 2: superior rectus recession alone (3 to 5 mm)

Length of follow-up:

Planned: not reported

Actual: mean 19 months (range 5 to 84 months) including randomized participants and non-randomized participants not included in this review; not reported by groups

Outcomes

Main outcomes: proportions of participants who were cured, improved, and unchanged; cure defined as latency or elimination of hyperdeviation; improved defined as intermittent but seldom noted or remained manifest but decreased in size and cosmetically acceptable angle; the study did not distinguish between primary and secondary outcomes

Adverse events: reported numbers of participants with postoperative hypotropia in the primary position and moderate limitation of elevation

Intervals at which outcomes assessed: not reported; data reported only at "last follow-up," which varied among participants

Notes

Publication type: journal article

Trial registration: not reported

Study period: not reported

Funding source: not reported

Disclosures of interest: not reported

Subgroup analyses: none reported

Contact with trial investigators: JMH contacted Dr von Noorden, but no additional study information available

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot reported
Allocation concealment (selection bias)Unclear riskNot reported
Masking of participants and personnel (performance bias)Unclear riskSurgeons cannot be masked. Whether participants were masked was not reported
Masking of outcome assessors (detection bias)Unclear riskNot reported
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNo report of incomplete data, but outcomes assessed at varying time points following surgery
Selective reporting (reporting bias)Unclear riskThe protocol was not available
Other biasUnclear riskNone evident

Quinn 2000

Methods

Study design: parallel-group RCT

Number randomized: 30 total participants (51 eyes); 15 participants (25 eyes) in the combined anteriorization with resection group and 15 participants (26 eyes) in the standard anteriorization group

Exclusions after randomization: none reported

Losses to follow-up: unclear; minimum of 4 months follow-up for all participants, but data analyzed at what seems to be the last available follow-up for each participant (i.e. no designated outcome time point at which to assess any loss to follow-up)

Number analyzed: 30 total participants (51 eyes); 15 participants (25 eyes) in the combined anteriorization with resection group and 15 participants (26 eyes) in the standard anteriorization group

Unit of analysis: the eye (21 participants had both eyes analyzed)

How was missing data handled?: no missing data reported for primary and secondary outcomes

Power calculation: yes; estimated a sample size of 15 participants in each group would have 80% power to detect an effect size of 0.75 when using a paired (before vs after surgery) comparison that included a 2-sided 5% significance level

Other study design details: none

Participants

Country: Canada

Age: mean 6 years (range 19 months to 15 years) overall; mean 6 years (range 20 months to 12 years) in the combined anteriorization with resection group and mean 6 years (range 19 months to 15 years) in the standard anteriorization group

Gender: 17 (57%) boys and 13 (43%) girls; 9 (60%) boys and 6 (40%) girls in the combined anteriorization with resection group and 8 (53%) boys and 7 (47%) girls in the standard anteriorization group

Inclusion criteria: DVD with or without inferior oblique overaction; DVD at least 5 pd in at least 1 eye and requiring surgery; ages 6 months to 18 years

Exclusion criteria: paretic or restrictive strabismus; prior oblique or vertical muscle surgery; prior or planned vertical off-set of horizontal recti; systemic disorder or syndrome that could affect alignment

Equivalence of baseline characteristics: yes, the 2 groups had similar age, gender, alignment complications, and reoperation rates

Interventions

Intervention 1: combined anteriorization of the inferior oblique muscle with resection (7 mm)

Intervention 2: standard anteriorization of the inferior oblique muscle

Length of follow-up:

Planned: minimum of 4 months

Actual: not reported overall; mean 25.2 months (range 4 to 42 months) in the combined anteriorization with resection group and mean 15.5 months (range 6 to 39 months) in the standard anteriorization group

Outcomes

Main outcomes: magnitude of postoperative DVD in prism diopters, grade of inferior oblique overaction, grade of superior oblique overaction, presence or absence of a true hypotropia in primary; the study did not distinguish between primary and secondary outcomes

Adverse events: reported numbers of participants with postoperative hypotropia in the primary position and mild limitation of elevation

Intervals at which outcomes assessed: not reported; data reported only at "last follow-up," which varied among participants

Notes

Publication type: journal article

Trial registration: not reported

Study period: March 1993 to January 1996

Funding source: "Dr Quinn’s fellowship was supported in part by a grant from the Glaucoma Trust of the Ophthalmological Society of New Zealand"

Disclosures of interest: not reported

Subgroup analyses: 1) assessed the magnitude of postoperative DVD in participants with little or no inferior oblique overaction preoperatively and in those with inferior oblique overaction of +1 or more; 2) assessed outcomes in participants with large angle DVD preoperatively (15 pd or more)

Contact with trial investigators: none

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot reported
Allocation concealment (selection bias)Unclear riskNot reported
Masking of participants and personnel (performance bias)Unclear riskSurgeons cannot be masked. Whether participants were masked was not reported
Masking of outcome assessors (detection bias)High risk13 of 15 participants (87%) in the combined anteriorization with resection group and 4 of 15 participants (27%) in the standard anteriorization group had masked outcome assessments. The disproportionately lower proportion of masked outcomes in the standard anteriorization group was taken to constitute a high risk of bias
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNo report of incomplete data, but outcomes assessed at varying time points following surgery
Selective reporting (reporting bias)Unclear riskThe protocol was not available
Other biasUnclear riskNone evident

Uncovska 2003

Methods

Study design: parallel-group RCT

Number randomized: 41 total participants (73 eyes); 22 participants (35 eyes) in the anteriorization group and 19 participants (38 eyes) in the myectomy group

Exclusions after randomization: none reported

Losses to follow-up: unclear; minimum of 7 months follow-up for all participants, but data analyzed at what seems to be the last available follow-up for each participant (i.e. no designated outcome time point at which to assess any loss to follow-up)

Number analyzed: 41 total participants (73 eyes); 22 participants (35 eyes) in the anteriorization group and 19 participants (38 eyes) in the myectomy group

Unit of analysis: the eye (32 participants had both eyes analyzed)

How was missing data handled?: no missing data reported for primary and secondary outcomes

Power calculation: none reported

Other study design details: none

Participants

Country: Czech Republic

Age: mean 6 years (range 13 months to 15 years); mean 5 years (range 13 months to 14 years) in the anteriorization group and mean 6 years (range 19 months to 15 years) in the myectomy group

Gender: not reported

Inclusion criteria: DVD with or without inferior oblique overaction, but any preoperative inferior oblique overaction was required to be concomitant; manifest or latent DVD at least 5 pd in one or both eyes and requiring surgery; ages 6 months to 15 years

Exclusion criteria: paretic or restrictive strabismus; prior oblique or vertical muscle surgery; prior vertical off-set of horizontal recti; systemic disorder or syndrome that could affect alignment

Equivalence of baseline characteristics: unclear; some baseline characteristics reported, but no formal comparison made; all participants had previous or planned horizontal muscle surgery

Interventions

Intervention 1: anteriorization of the inferior oblique muscle

Intervention 2: myectomy of the inferior oblique muscle

Length of follow-up:

Planned: not reported

Actual: not reported overall; mean 19.6 months (range 7 to 43 months) in the anteriorization group and mean 26.5 months (range 8 to 52 months) in the myectomy group

Outcomes

Main outcomes: for eyes with more than 15 pd of DVD preoperatively, outcomes were classified as either excellent (0 to 4 pd), good (5 to 9 pd), fair (10 to 14 pd), or poor (> 14 pd) and compared between treatment groups; mean magnitude of postoperative DVD in prism diopters compared with mean preoperative DVD within each group (groups not compared); grade of postoperative inferior oblique overaction compared pre- versus postoperatively; the study did not distinguish between primary and secondary outcomes

Adverse events: reported numbers of participants with limitation of elevation and needing repeat surgeries

Intervals at which outcomes assessed: not reported; data reported only at "last follow-up," which varied among participants

Notes

Publication type: journal article

Trial registration: not reported

Study period: 1994 to 1999

Funding source: not reported

Disclosures of interest: not reported

Subgroup analyses: 1) assessed the effect of interventions on postoperative inferior oblique overaction; 2) assessed outcomes in participants with large angle DVD preoperatively (15 pd or more)

Contact with trial investigators: email contact was attempted but unsuccessful

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot reported
Allocation concealment (selection bias)Unclear riskNot reported
Masking of participants and personnel (performance bias)Unclear riskSurgeons cannot be masked. Whether participants were masked was not reported
Masking of outcome assessors (detection bias)Unclear riskMasked assessment was performed by an orthoptist when possible, but the actual number of masked exams performed in each group was not reported
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNo report of incomplete data, but outcomes assessed at varying time points following surgery
Selective reporting (reporting bias)Unclear riskThe protocol was not available
Other biasUnclear riskNone evident

Vodickova 2008

  1. a

    DVD: dissociated vertical deviation
    pd: prism diopters
    RCT: randomized controlled trial

Methods

The full study was written in Czech but was translated into English for data abstraction

Study design: parallel-group RCT

Number randomized: 46 total participants (82 eyes); 27 participants (44 eyes) in the anteriorization group and 19 participants (38 eyes) in the myectomy group

Exclusions after randomization: none reported

Losses to follow-up: unclear; minimum of 7 months follow-up for all participants, but data analyzed at what seems to be the last available follow-up for each participant (i.e. no designated outcome time point at which to assess any loss to follow-up)

Number analyzed: 46 total participants (82 eyes); 27 participants (44 eyes) in the anteriorization group and 19 participants (38 eyes) in the myectomy group

Unit of analysis: the eye (36 participants had both eyes analyzed)

How was missing data handled?: no missing data reported for primary and secondary outcomes

Power calculation: none reported

Other study design details: This study is suspect for being a reproduction of the Uncovska 2003 study, since the number of participants, baseline characteristics, and postoperative alignment for the myectomy group are exactly the same in the 2 studies (both studies were conducted in the Czech Republic). The only difference between Uncovska 2003 and Vodickova 2008 regarding the study population is that the Vodickova 2008 study had an additional 5 participants in the anteriorization group. The other difference is that Vodickova 2008 reported slightly longer follow-up in the myectomy group. There is no reference by the Vodickova 2008 study authors to the Uncovska 2003 study; the Vodickova 2008 study is reported as an original RCT. We consider it highly unlikely that an independent RCT would produce participants with exactly the same demographics and outcomes in 1 of the treatment groups as those in a previous RCT

Participants

Country: Czech Republic

Age: mean 6 years (range 13 months to 15 years); mean 5 years (range 13 months to 14 years) in the anteriorization group and mean 6 years (range 19 months to 15 years) in the myectomy group

Gender: not reported

Inclusion criteria: DVD with or without inferior oblique overaction; manifest or latent DVD at least 5 pd in one or both eyes and requiring surgery; ages 6 months to 15 years

Exclusion criteria: paretic or restrictive strabismus; prior oblique or vertical muscle surgery; systemic disorder or syndrome that could affect extraocular muscle function

Equivalence of baseline characteristics: unclear; some baseline characteristics reported, but no formal comparison made

Interventions

Intervention 1: anteriorization of the inferior oblique muscle

Intervention 2: myectomy of the inferior oblique muscle

Length of follow-up:

Planned: not reported

Actual: not reported overall; mean 29.6 months (range 7 to 53 months) in the anteriorization group and mean 36.5 months (range 8 to 72 months) in the myectomy group

Outcomes

Main outcomes: for eyes with more than 15 pd of DVD preoperatively, outcomes were classified as either excellent (0 to 4 pd), good (5 to 9 pd), fair (10 to 14 pd), or poor (> 14 pd) and compared between treatment groups; median magnitude of postoperative DVD in prism diopters compared with median preoperative DVD within each group (groups not compared); grade of postoperative inferior oblique overaction compared pre- versus postoperatively; the study did not distinguish between primary and secondary outcomes

Adverse events: reported numbers of participants with limitation of elevation and needing repeat surgeries

Intervals at which outcomes assessed: not reported; data reported only at "last follow-up," which varied among participants

Notes

Publication type: journal article

Trial registration: not reported

Study period: 1995 to 2004

Funding source: not reported

Disclosures of interest: not reported

Subgroup analyses: assessed outcomes in participants with large angle DVD preoperatively (15 pd or more)

Contact with trial investigators: email contact was made, but no additional study information available

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot reported
Allocation concealment (selection bias)Unclear riskNot reported
Masking of participants and personnel (performance bias)Unclear riskSurgeons cannot be masked. Whether participants were masked was not reported
Masking of outcome assessors (detection bias)Unclear riskNot reported
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNo report of incomplete data, but outcomes assessed at varying time points following surgery
Selective reporting (reporting bias)Unclear riskThe protocol was not available
Other biasHigh riskPlease refer to notes in the above table outlining study characteristics. We considered there to be a high risk that the myectomy group in this study is the same as that reported in another RCT, and that the participants in the anteriorization group include the same participants from the other RCT (Uncovska 2003)

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
  1. a

    ARVO: Association for Research in Vision and Ophthalmology
    DVD: dissociated vertical deviation
    RCT: randomized controlled trial

Apers 1981Description of methods to assess DVD. Not a RCT on interventions for DVD
Castellanos-Bracamontes 1997Description of a surgical technique. Not a RCT on interventions for DVD
Escalante-Razo 2002This is an ARVO abstract, and no corresponding full-length publication was found. We contacted the study investigator but did not receive a reply. Given unclear inclusion criteria of participants, we chose to exclude this study
Kratz 1989Comparison of two surgical approaches, but treatment group allocation was not randomized
Velez 1993Different procedures performed on different patient subgroups. Not a randomized treatment comparison

Ancillary