Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Triaden som bevis för skakvåld - En undersökning av domstolarnas bevisvärdering efter NJA 2014 s. 699

Fagerberg, Emma LU (2020) JURM02 20201
Department of Law
Faculty of Law
Abstract (Swedish)
Länge ansågs en triad av symtom, bestående av blödning under den hårda hjärnhinnan, ögonbottenblödning och hjärnpåverkan, innebära att ett barn hade blivit skakat. Triaden uppfattades även av svenska domstolar som stark bevisning för att ett barn blivit skakat och det finns flera fall där den tilltalade dömts med enbart triaden som bevisning. Anledningen till detta är att de experter som hördes i målen var överens om att symtomen kunde förklara uppkomstsättet. Under början av 2010-talet började dock triaden som bevis för skakvåld att ifrågasättas och det bildades två läger inom forskarvärlden. I oktober 2014 kom Högsta domstolen med ett avgörande, NJA 2014 s. 699, som helt gick emot den praxis som tidigare funnits rörande skakvåld. I... (More)
Länge ansågs en triad av symtom, bestående av blödning under den hårda hjärnhinnan, ögonbottenblödning och hjärnpåverkan, innebära att ett barn hade blivit skakat. Triaden uppfattades även av svenska domstolar som stark bevisning för att ett barn blivit skakat och det finns flera fall där den tilltalade dömts med enbart triaden som bevisning. Anledningen till detta är att de experter som hördes i målen var överens om att symtomen kunde förklara uppkomstsättet. Under början av 2010-talet började dock triaden som bevis för skakvåld att ifrågasättas och det bildades två läger inom forskarvärlden. I oktober 2014 kom Högsta domstolen med ett avgörande, NJA 2014 s. 699, som helt gick emot den praxis som tidigare funnits rörande skakvåld. I rättsfallet kom HD fram till, med hänvisning till en systematisk litteratur-översikt som SBU höll på att genomföra om just med vilken säkerhet triaden kan förklaras av skakvåld, att triaden i princip inte ensamt kunde bevisa att skakvåld ägt rum. HD uttalade att det krävs att slutsatsen vilar på en vetenskaplig ståndpunkt som det finns mycket starka belägg för samt att det i praktiken måste vara uteslutet att det finns någon alternativ förklaring. Även SBU kom i sin rapport fram till att det fanns otillräckligt vetenskapligt stöd för att bedöma den diagnostiska träffsäkerheten av triaden för att identifiera skakvåld samt att det fanns begränsat vetenskapligt underlag för att triaden och därmed dess delkomponenter kan förekomma vid skakvåld.

Det huvudsakliga syftet med denna uppsats är att se hur skakvåld och triaden behandlas i domstolarna idag, efter NJA 2014 s. 699, samt att undersöka vilken betydelse och vilka konsekvenser HD:s uttalanden har fått. I uppsatsen diskuteras även om domstolarnas bevisvärdering är acceptabel eller om den kan förbättras.

Av den rättsfallsundersökning som görs i uppsatsen kan slutsatsen dras att det efter NJA-fallet i princip bara meddelas slutligt friande domar i mål om skakvåld. Tillsammans med en bayesiansk utvärdering av NJA 2014 s. 699 och ett senare hovrättsfall tyder detta på att triaden i princip inte får någon betydelse alls numera. Det går även att dra slutsatsen att det verkar finnas en svårighet i att kombinera medicinsk bevisning bestående av triaden med annan typ av bevisning samt att domstolen slentrianmässigt behandlar triaden som en helhet och inte ser till omständigheterna i det specifika fallet. Utifrån principen om fri bevisföring och bevisvärdering skulle denna utveckling av många anses icke acceptabel, då det verkar som att domstolarna nu inte gör en självständig och fri värdering av bevisningen. HD:s uttalanden har uppenbarligen fått stora konsekvenser när det kommer till hur de lägre instanserna dömer, något det går att argumentera för att HD borde ha förutsett. Vad gäller vilka intressen som får företräde i skakvåldsfall kan slutsatsen dras att barnets intresse av att inte bli utsatt för våld i sitt eget hem har hamnat i skymundan till förmån för intresset av att vårdnadshavaren inte ska bli oskyldigt dömd.

Då det största problemet verkar vara att domstolarna inte klarar av att hantera bevisningen på ett självständigt sätt skulle en lösning kunna vara att domstolarna lär sig att använda bayesianska nätverk. Genom dessa skapas ett medvetande om vilka överväganden och antaganden som görs vid bevis-värderingen. Även om triaden oftast inte ensam kan nå upp till beviskravet i brottmål måste den ges en viss beviskraft som kombinerat med annan bevisning kan leda till fällande domar. Det behöver ske en förändring i det sätt på vilket domstolarna idag på ren rutin bortser från triaden som bevis för skakvåld. (Less)
Abstract
For a long time, a triad of symptoms, namely subdural hematoma, retinal bleedings and brain impacts, were considered an indication that a child had been shaken (the shaken baby syndrome). The triad was perceived as strong evidence by Swedish courts, demonstrated by several cases where the defendants were convicted based exclusively on the existence of this triad. The reason for this was that expert testimonies established a consensus that a finding of these symptoms was also indicative of the cause of the trauma in question. However, in the beginning of the 2010s, scientists began to question the validity of the triad as evidence of a child being shaken, dividing the scientific community. In October 2014 The Swedish Supreme Court (HD)... (More)
For a long time, a triad of symptoms, namely subdural hematoma, retinal bleedings and brain impacts, were considered an indication that a child had been shaken (the shaken baby syndrome). The triad was perceived as strong evidence by Swedish courts, demonstrated by several cases where the defendants were convicted based exclusively on the existence of this triad. The reason for this was that expert testimonies established a consensus that a finding of these symptoms was also indicative of the cause of the trauma in question. However, in the beginning of the 2010s, scientists began to question the validity of the triad as evidence of a child being shaken, dividing the scientific community. In October 2014 The Swedish Supreme Court (HD) departed from precedent in decision NJA 2014 p. 699 with respect to such evidentiary standards. In its judgment, HD found that, in principle, the triad alone cannot prove violent shaking. In this finding, the court referred to a systematic literature review conducted by SBU on the certainty of the triad as evidence of violent shaking. HD found that it is necessary that the assessment of whether or not the violence has occurred be based on a very well-founded scientific position and that no alternative explanation for the injuries exist. In its report, SBU concluded that there was insufficient scientific support for assessing the diagnostic accuracy of the triad as a method of identifying violent shaking and, further, that there was limited scientific evidence that the triad and its subcomponents occurs in violent shaking.

The main purpose of this thesis is twofold: first, to examine how violent shaking and the triad are dealt with in the Swedish courts following NJA 2014 p. 699 and second, to analyze the significance and consequences of HD's findings. The thesis also discusses whether the evidentiary assessments carried out by the courts is acceptable or if they can be improved.

Following an analysis of the available jurisprudence, this thesis concludes that after the NJA-case, court assessments of violent shaking cases overwhelmingly find in favor of the defendant. The analysis paired with a Bayesian evaluation of both NJA 2014 p. 699 and a later case indicates that the triad has essentially lost its significance as evidence in violent shaking cases. The courts also appear to be challenged in combining medical evidence (the triad) with other types of evidence. Moreover, the court casually treats the triad as a unit and does not consider the specific circumstances of the individual case. This development is unacceptable because the systematic failure of the courts to undertake a free and independent evaluation of the evidence at hand, in contravention with the principle of free submission and evaluation of evidence. HD's findings have clearly had significant consequences when it comes to how lower courts rule, which HD should arguably have foreseen. With regard to whose interests take precedence in violent shaking cases, it appears that avoiding wrongful conviction of parents is currently being prioritized over the interest of the child in not being subjected to domestic violence.

Since the main problem appears to be that the courts are unable to make independent assessments of evidence in violent shaking cases, a solution may be that the courts learn to use Bayesian networks. These networks create an awareness among judges of what considerations and assumptions are typically made in the evaluation of evidence in violent shaking cases. While existence of the triad alone is generally not considered sufficient to meet the evidentiary requirement in criminal cases, it cannot be wholly dismissed. The triad must be taken into consideration as evidence which, when considered in combination with other evidence, can lead to convicting judgments. The systematic dismissal of the triad as evidence in violent shaking cases currently carried out by the courts calls for a change in approach to the assessment of evidence in these cases. (Less)
Please use this url to cite or link to this publication:
author
Fagerberg, Emma LU
supervisor
organization
alternative title
The Triad as Evidence for Shaken Baby Syndrome - a Study of the Courts' Evaluation of Evidence after NJA 2014 p. 699
course
JURM02 20201
year
type
H3 - Professional qualifications (4 Years - )
subject
keywords
allmän rättslära, fri bevisvärdering, bevisvärdering, beviskraft, bayesiansk bevisvärderingsmetod, skakvåld, triaden, shaken baby syndrome, abusive head trauma
language
Swedish
id
9010514
date added to LUP
2020-06-16 10:55:36
date last changed
2020-06-16 10:55:36
@misc{9010514,
  abstract     = {{For a long time, a triad of symptoms, namely subdural hematoma, retinal bleedings and brain impacts, were considered an indication that a child had been shaken (the shaken baby syndrome). The triad was perceived as strong evidence by Swedish courts, demonstrated by several cases where the defendants were convicted based exclusively on the existence of this triad. The reason for this was that expert testimonies established a consensus that a finding of these symptoms was also indicative of the cause of the trauma in question. However, in the beginning of the 2010s, scientists began to question the validity of the triad as evidence of a child being shaken, dividing the scientific community. In October 2014 The Swedish Supreme Court (HD) departed from precedent in decision NJA 2014 p. 699 with respect to such evidentiary standards. In its judgment, HD found that, in principle, the triad alone cannot prove violent shaking. In this finding, the court referred to a systematic literature review conducted by SBU on the certainty of the triad as evidence of violent shaking. HD found that it is necessary that the assessment of whether or not the violence has occurred be based on a very well-founded scientific position and that no alternative explanation for the injuries exist. In its report, SBU concluded that there was insufficient scientific support for assessing the diagnostic accuracy of the triad as a method of identifying violent shaking and, further, that there was limited scientific evidence that the triad and its subcomponents occurs in violent shaking.

The main purpose of this thesis is twofold: first, to examine how violent shaking and the triad are dealt with in the Swedish courts following NJA 2014 p. 699 and second, to analyze the significance and consequences of HD's findings. The thesis also discusses whether the evidentiary assessments carried out by the courts is acceptable or if they can be improved. 

Following an analysis of the available jurisprudence, this thesis concludes that after the NJA-case, court assessments of violent shaking cases overwhelmingly find in favor of the defendant. The analysis paired with a Bayesian evaluation of both NJA 2014 p. 699 and a later case indicates that the triad has essentially lost its significance as evidence in violent shaking cases. The courts also appear to be challenged in combining medical evidence (the triad) with other types of evidence. Moreover, the court casually treats the triad as a unit and does not consider the specific circumstances of the individual case. This development is unacceptable because the systematic failure of the courts to undertake a free and independent evaluation of the evidence at hand, in contravention with the principle of free submission and evaluation of evidence. HD's findings have clearly had significant consequences when it comes to how lower courts rule, which HD should arguably have foreseen. With regard to whose interests take precedence in violent shaking cases, it appears that avoiding wrongful conviction of parents is currently being prioritized over the interest of the child in not being subjected to domestic violence.

Since the main problem appears to be that the courts are unable to make independent assessments of evidence in violent shaking cases, a solution may be that the courts learn to use Bayesian networks. These networks create an awareness among judges of what considerations and assumptions are typically made in the evaluation of evidence in violent shaking cases. While existence of the triad alone is generally not considered sufficient to meet the evidentiary requirement in criminal cases, it cannot be wholly dismissed. The triad must be taken into consideration as evidence which, when considered in combination with other evidence, can lead to convicting judgments. The systematic dismissal of the triad as evidence in violent shaking cases currently carried out by the courts calls for a change in approach to the assessment of evidence in these cases.}},
  author       = {{Fagerberg, Emma}},
  language     = {{swe}},
  note         = {{Student Paper}},
  title        = {{Triaden som bevis för skakvåld - En undersökning av domstolarnas bevisvärdering efter NJA 2014 s. 699}},
  year         = {{2020}},
}