Decomposition scoring method
Table 1 shows the developed decomposition scoring method. This method was developed after an extensive literature study. Each stage was assigned a score. Each body part had a maximum score of six and the TDS ranges between 3 and 18.
The first stage is called no visible signs. An individual just passed away and there are no decomposition signs visible yet. The second stage is where the livor mortis, rigour mortis, algor mortis and vibices occur. Drying of the body can lead to tache noir (when the eyes are open) and brownish discoloration starting at the edges (nose, ears, lips, fingers and toes) expands to the rest of the body . Cloudy eyes can occur as soon as 10 min post-mortem if the eyes are open. If the eyes are shut, the cornea becomes cloudy after 24 h . Tache noir can occur as soon as 1–2 h post-mortem . The appearance of livor mortis is different for each human remain and can range between 15 min and many hours [37, 38]. It is also possible that it does not occur at all . Rigour mortis is commonly detected between 3 and 6 hours. The face shows stiffening between 1 and 4 hours after death, and the limbs show stiffening between 3 and 6 hours after death [23, 28]. According to Galloway et al., livor mortis is followed by drying of the extremities . Since the order in which the phenomena occur is variable, they were not given their own score.
The putrefaction, caused by autolysis, is not visible in stage two and starts in stage three with grey to green discoloration caused by the bacteria . The same bacteria produce gasses which give the face and trunk a swollen appearance (bloating), starting in the areas with low turgor (eyelids, scrotum). The internal pressure caused by these gasses causes the tongue to protrude from the mouth, the abdomen to bloat and purging of putrefaction fluids out of the natural body openings (ears, nose, mouth and rectum) [25, 39]. These fluids leave a strong odour behind [25, 39]. The build-up pressure in the abdomen causes the blood to be pushed into the head, which causes the brown to black discoloration . Also in this stage, skin blisters filled with decomposition fluids form and skin lets loose due to hydrolytic enzymes (slippage). Marbling is the result of gasses reacting with haemoglobin producing a greenish pigment in the veins [23, 39]. Further into this stage, the skin turns dark green to black [23, 38]. Phenomena as bloating, skin blisters and marbling can occur at the same time or one before the other in random order and rapidly after each other [10, 24]. Researchers disagree about the order in which the phenomena occur. Some researchers found bloating first and blistering and marbling later . Some researchers found blistering and marbling first and later bloating [38, 40,41,42]. Some researchers found marbling first and later blisters and bloating [37, 43]. For this reason, these phenomena cannot be put in a rigid order.
Stage four starts when the putrefaction reaches an end and gasses are released and the tissue caves in, because tissues do not have the strength or the capability anymore, due to autolysis, to uphold connections. This gives the eyes, throat and abdominal cavity a hollow appearance . Due to dehydration and draining of bodily fluids, the skin dries and gets a leathery appearance with discoloration and darkening of the skin. It becomes hard and contracts tightly around the bones . The leathery skin forms a shell over the body and could protect the remaining decomposing tissue underneath [8, 24]. Also, partial skeletonization is visible, which means less than 50% bone is visible in that body area (face/neck, trunk or limbs). The joints are still together. According to Galloway et al., it is first caving in of the flesh followed by dehydration and the skin turning leathery. This in turn is followed by bone exposure (less than 50%) .
Stage five contains gross skeletonization which means most of the skin is gone and no soft tissue is left. More than 50% of bone is visible in that area which means either in the jaw, cheekbones, skull or a combination for the face/neck area. In the trunk, it could be the ribcage, the thoracic or lumbar vertebrae, the pelvic bones or a combination. In the limbs, it is a combination of the humerus, ulna, radius, wrist bones, the bones of the hand, the femur, the tibia or fibula or the bones of the feet. Some joints can be disarticulated such as the jaw, the cervical, thoracic or lumbar vertebrae, the ribcage or pelvic joints. In the limbs, it could be the shoulder, elbow, wrist, knee and ankle.
The last stage contains complete skeletonization (stage six) where only bones, cartilage, hair and sometimes a bit of skin are all that are left from the human remains [25, 39]. The skeletonization can be considered complete, when all soft tissue is removed .
The Book of Reference is a visual resource containing photographic examples of all the decomposition phenomena with their corresponding scores of the stage they are in. It contains a total of 84 specifically selected photographs. The photographs are placed in a chronological order, following the point-based system of the decomposition scoring method. Before each region is visually explained, the decomposition scoring method is shown in a table. Figures 2 and 3 are examples and part of the Book of Reference. Since stage 3.2 of both FDS and BDS and stage 3.1 of the LDS have multiple phenomena and all the phenomena have to be present in the Book of Reference, these stadia were further divided into subgroups, hence the reason for 3.2.3 in Fig. 2.
Twelve participants scored 45 test photographs. Table 2 shows the Fleiss kappa values. The lowest kappa score was 0.74 for the facial decomposition scored by the medical students (moderate agreement). The highest kappa score was 0.97 for the body decomposition score, also scored by the medical students (high agreement). In total, there are two values showing moderate agreement (FDS scored by students and LDS scored by forensic physicians) and seven values showing a high agreement .
Estimating the post-mortem interval using the decomposition scoring method and ADD
Of the 91 cases with a known PMI, 28 individuals were female (30.8%) and 63 were male (69.2%) (see Appendix 1 for an overview of the characteristics, available online only). The age varied from 20 to 92 with a mean value of 54 (standard deviation (SD) ± 18). Most of the individuals died indoors (n = 79, 86.8%) and 12 individuals were found outside (13.2%). Twenty-nine cases were found in the Spring (31.9%), 19 in the Summer (20.9%), 21 in the Autumn (23.1%) and 22 in the Winter (24.2%).
The decomposition scores varied from score 3 to score 12. The median TDS is six, with an interquartile range (IQR) of six to nine. A TDS of six is by far the most common score (44%), which means a lot of individuals showed signs of livor mortis or rigour mortis. The second most common score is nine (18.7%). These individuals were in the putrefaction stage.
Of the 79 indoor cases, two individuals were found in a shed and one individual was found in a garage. Here, it was definitely unknown what the temperature was. It could not be decided if it was a normal inside temperature or a temperature close to the outside temperature. The ADD ranged from 4 to 1080 (n = 76). The median ADD is 28.5 with an IQR of 18 to 90. Of the 76 cases, 59 cases had an estimated mean temperature of 18 °C (77.6%). For the outdoor cases (n = 12), the ADD ranged from 3 to 512 with a median of 14.5 and an IQR of 8.3 to 19. In the 12 cases where the individuals were found outdoors, the weather station was as close as 0 km and as far as 29 with an IQR of 14 to 19 km .
The correlation between the decomposition scores and PMI was the lowest between FDS and PMI (r = 0.697, p = 0.000, n = 91). The correlation between LDS and PMI was second lowest (r = 0.774, p = 0.000, n = 91) and the second highest correlation was between BDS and PMI (r = 0.801, p = 0.000, n = 91) (see Fig. 4). The highest correlation was found between TDS and PMI (r = 0.812, p = 0.000, n = 91). Even though the correlation between FDS and PMI is the lowest, it is still considered to be a strong correlation. The remaining three correlations are considered to be ‘very strong’ .
The relationship between PMI and TDS and ADD and TDS is exponential (curvilinear). The PMI and ADD were log transformed to achieve a linear relationship. After separating the indoor and outdoor cases, the following four formulas could be made with their associated standard errors. Starting with the indoor cases (n = 79), the TDS significantly estimated the PMI (β = 0.82, p = 0.000). The R
2 indicated that 67% of the variation in the PMI was estimated by the TDS. The developed formula is PMI = 10^(− 1.18 + 0.22·TDS) (see Fig. 5). The associated standard error of the regression in untransformed form (non-logged PMI) is 1.6 days. When using the TDS to estimate the ADD, the R
2 decreased to 66% (β = 0.81, p = 0.000). The developed formula is ADD = 10^(− 0.05 + 0.23·TDS) (see Fig. 6). The associated standard error of the regression in untransformed form (non-logged ADD) is 29.6 ADD. There were 12 cases outdoors. The TDS significantly estimated the PMI (β = 0.90, p = 0.000). The R
2 indicated that 80% of the variation in the PMI was estimated by the TDS. The developed formula is PMI = 10^(− 0.93 + 0.18·TDS) (see Fig. 7). The associated standard error of the regression in untransformed form (non-logged PMI) is 2.9 days. When using the TDS to estimate the ADD, the R
2 decreased to 56% (β = 0.75, p = 0.005). The developed formula is ADD = 10^(0.03 + 0.19·TDS) (see Fig. 8). The associated standard error of the regression in untransformed form (non-logged ADD) is 52.1 ADD. The standard errors were calculated with the t distribution (see Table 3).