Plagiarism in the PhD thesis of Prof. Cheung Lim-kwong

Prof. Cheung Lim-kwong was the Associate Dean of Research, the former Chairman of the Faculty Higher Degrees Committee at the HKU Faculty of Dentistry and a member of the Board of Graduate Studies. Prof. Cheung's PhD thesis titled The temporalis myofascial flap in maxillofacial reconstruction: vascular anatomy and healing was previously available for downloading at HKU Scholars hub at http://hub.hku.hk/bitstream/10722/128702/1/FullText.pdf . However following the exposure of the plagiarism scandal HKU has blocked online access to this thesis. 

A side by side comparison reveals the degree of plagiarism. This may be only the tip of the iceberg. Most journal articles during the 1990's & 80's are only available in the hard copy format. Prof. Cheung has also included in the PhD, paragraphs copied from some text books. So it is quite difficult to locate other potential sources he may have plagiarized.

Page No.	Prof. Lim Cheung's PhD	Original Text	Original Source
5	The temporalis muscle inserts into the coronoid process and the anterior aspect of the ramus of the mandible.	The muscle inserts into the coronoid process and anterior aspect of the vertical ramus of the mandible.	O. Antonyshyn, J. S. Gruss and B. D. Birt. Versatility of temporal muscle and fascial flaps.  British Journal of Plastic Surgery (1988), 41,118-131
7	The superficial temporal fascia is a thin, highly vascular layer of moderately dense connective tissue. It lies immediately deep to the hair follicles and the subdermal layer of fibrofatty tissue.	The temporoparietal fascia is a thin layer of moderately dense connective tissue, which lies immediately deep to the hair follicles and subdermal fibro-fatty tissue.
7	....  is part of the superficial musculoaponeurotic system and is continuous with the fascia over the parotid gland below and with the galea aponeurosis above.	is part of the subcutaneous musculo-aponeurotic system, and is continuous with the galea superiorly and the superficial musculo-aponeurotic system inferiorly
7	The superficial temporal fascia is separated from the deep temporal fascia by a distinct layer of loose areolar tissue.	It is separated from the underlying deep temporal fascia and temporalis muscle by a loose avascular plane.
7	The deep temporal fascia is a dense, uniform fascial layer which completely invests the outer aspect of the temporalis muscle. At the periphery of the muscle all round, the fascia fuses with the pericranium at the superior temporal line.	.....the deep temporal fascia. This is a dense fascial layer which completely invests the outer surface of the muscle from the upper edge of the zygoma, and peripherally fuses with pericranium at the superior temporal line of the skull.
10	obliterating the dead space following orbital exenteration (Naquin 1956, Webster 1957, Reese 1958, Reese & Jones 1961), and	obliteration of dead space following orbital exenteration (Naquin, 1956; Reese and Jones, 1961; Deitch and
18	Considerable interest has developed in the combined use of the temporalis muscle and its underlying periosteum to provide a single vascularized unit with bone forming capacity.	Most recently, considerable interest has developed in the combined use of temporalis muscle and the underlying pericranium to provide a single vascularised unit with bone-forming capacity.
18	The osteogenic potential of the temporal musculoperiosteal flaps has been confirmed experimentally in young animals (Hauben & van der Meulen 1984).	The osteogenic potential of temporal musculoperiosteal flaps has been confirmed experimentally in young animals (Hauben and van der Meulen, 1984)
18	Its immediate clinical application, however, is restricted by the possible deleterious effects that early masticatory muscle transposition may have on subsequent craniofacial growth (Hohl 1983).	Its immediate clinical application is, however, restricted by the possible deleterious effects that early masticatory muscle transposition may have on subsequent craniofacial growth (Hohl, 1983).
72	..the anterior and posterior deep temporal vessels supply the anterior and posterior part of the muscle respectively...	...the anterior and posterior deep temporal arteries, supply the anterior and posterior portions of the muscle, respectively (Mathes and Nahai, 1982).
9	Lexer (1908) and Rosenthal (1916) utilized the TMF to reanimate the eyelid following paralysis of the facial nerve.	Lexer (1908) and Rosenthal (1916) utilized the TMF to reanimate the eyelid following paralysis of the facial nerve,	Clauser L, , Curioni C and Spanio S. The use of the temporalis muscle flap in facial and craniofacial reconstructive surgery. A review of 182 cases. Journal of Cranio-Maxillofacial Surgery Volume 23, Issue 4, August 1995, Pages 203-214
11	Ewers (1988) described a different technique of reconstruction after maxillectomy with the.....	Ewers (1988), reported a new technique of reconstruction of the
12	Colmenero et al. (1991) reported their experience of 26 TMFs....	Colmenero et al. (1991), reported their experience gained with 26 temporalis flaps....
12	...as a composite flap by combining it with cranial bone, coronoid process or a temporal skin island. Although major complications were not observed,	.....as a composite flap with cranial bone, coronoid process or skin island. Major complications were not observed.
12	total necrosis of the flap developed in 3 cases	,total necrosis of the TMF occurred in 3 cases
12 & 13	should be taken into consideration before deciding on more extensive reconstructive procedures.	should be taken into consideration before deciding on more extensive reconstructive procedures.
13	Van der Wal & Mulder (1992) reported another 4 cases of large palatal defects in cleft lip and palate	Van der Wal and Mulder (1992), reported closure of large palatal defects in four patients with congenital cleft palate.
38	the residual posterior part of the muscle transposed to fill the anterior temporal fossa,	The posterior part of the muscle was then advanced to fill in the depression in the temporalis fossa.
9	The first use of the temporalis myofascial flap (TMF) has been attributed to Golovine	the origin of the temporalis muscle flap which has been attributed to Golovine,	B. Speculand. The origin of the temporalis muscle flap. British Journal of Oral and Maxillofacial Surgery Volume 30, Issue 6, December 1992, Pages 390-392
9	Sir Harold Gillies in 1917 described a series of cases where the temporalis muscle was used as a transposition flap for deformities caused by the loss of the zygomatic bone,	Sir Harold Gillies in 1917 described a series of cases where temporalis muscle was used as a transposition flap for deformities caused by loss of the malar bone.
9	In a later paper (Gillies 1934), he described the use of TMF, tunnelled to	In a later paper (Gillics, 1934) he described the use of temporalis muscle and fascia tunnelled to
9	...either the corner of the mouth or the inner canthus of the eye, for facial reanimation.	...either the corner of the mouth or the inner canthus of the eye for reanimation.
5	The temporalis muscle inserts into the coronoid process and the anterior aspect of the ramus of the mandible.	...the temporalis msucle is inserted into the coronoid process and the anterior border of the ramus of the mandible...	Harn D, Shackelford LS. Further evaluation of the superficial and deep tendons of the human temporalis muscle. The Anatomical Record. , April 1982. Volume 202, Issue 4, pages 537–548
5	DuBrul (1980) pointed out that the tendinous attachment to the mandible can be divided into the superficial and deep tendons. The superficial tendon inserts into the anterior border of the coronoid process and the deep tendon inserts into the internal oblique line reaching down to the retromolar pad.	DuBrul (1980) presents further detail stating that there are superficial and deep tendons which the former inserts into the anterior border of the coronoid process and the latter inserts into the temporal crest (internal oblique line) reaching the area of the lower third molar into the retromolar pad.
54	The fact that glycogen decreases as the degree of keratosis increases suggests a role for glycogen in keratinization. The glycogen may serve as a source of energy for keratin synthesis.	The fact that glycogen content decreases as the degree of keratosis increases suggests a role for glycogen in keratinization. The glycogen may serve as a source of energy required for keratin synthesis	Silverman S, Barbosaa J, Kearnsa G. Ultrastructural and histochemical localization of glycogen in human normal and hyperkeratotic oral epithelium. Archives of Oral Biology. Volume 16, Issue 4, April 1971, Pages 423-432,
92	.....with the marginal cells at the advancing front being the active motile cells, while the cells behind the margins are passively dragged along	the cells at the margin of the moving sheet appeared to be actively motile while the cells behind (or above, in a stratified layer) the marginal cells were passively dragged along	Page 119. Wound healing : biochemical & clinical aspects / [edited by] I. Kelman Cohen, Robert F. Diegelmann, William J. Lindblad. Philadelphia : W.B. Saunders Co., c1992.
92	This mode of sheet movement, referred to as the sliding model of wound closure, has been demonstrated on epithelial cells in cell culture (Vaughan & Trinkaus 1966), in embryo (Bellairs 1963), and in corneal healing (Buck 1979).	This mode of sheet movement, referred to as the sliding model of wound closure, has been demonstrated directly for epithelial cells in tissue culture,36 for embryonic epithelial moveme nt,38 for amphibian wound closure,37 and for corneal wound closure.39	Page 119. Wound healing : biochemical & clinical aspects / [edited by] I. Kelman Cohen, Robert F. Diegelmann, William J. Lindblad. Philadelphia : W.B. Saunders Co., c1992.
92 & 93	It is more difficult to study mammalian cutaneous wound closure directly because of the thickness and opacity of the dermis. Moreover, the migrating epithelial sheet in the mammalian epidermis is multilayered and thus more complex than those systems illustrated in the sliding model.	It is much more difficult to study mammalian cutaneous wound closure directly because of the thickness and opacity of the dermis. Moreover, the migrating epithelial sheet of mammalian epidermis is multilayered and thus more complex than those systems illustrating the sliding model.	Page 119. Wound healing : biochemical & clinical aspects / [edited by] I. Kelman Cohen, Robert F. Diegelmann, William J. Lindblad. Philadelphia : W.B. Saunders Co., c1992.
93	For the repair of mammalian epidermis, Winter (1964) proposed the "leap frog model" of epidermal sheet movement. This model was deduced from morphological data at ultrastructural level, which suggested that cells at the migrating front adhere to the substrate......	For the repairing mammalian epidermis, Winter40 proposed the leap frog model of epidermal sheet movement (Fig. 7- 4). This model was deduced indirectly from ultrastructural morphological data that suggested that cells at the migrating front adhere to the substrate.........	Page 119. Wound healing : biochemical & clinical aspects / [edited by] I. Kelman Cohen, Robert F. Diegelmann, William J. Lindblad. Philadelphia : W.B. Saunders Co., c1992.
93	.......submarginal cells are thus conceived to crawl over the newly adherent basal cells in a "leap frog" fashion (Krawczyk 1971, Kuwabara et al. 1976, Gibbins 1978).	.......submarginal cells are conceived to crawl over the newly adherent basal cells in a "leapfrog" fashion	Page 119. Wound healing : biochemical & clinical aspects / [edited by] I. Kelman Cohen, Robert F. Diegelmann, William J. Lindblad. Philadelphia : W.B. Saunders Co., c1992.
93	What actually turns on the cellular machinery of movement, be it physical or chemical, is still not known.	what actually "turns on" the cellular machinery of movement, be it physical or chemical, is still not known.	Page 124. Wound healing : biochemical & clinical aspects / [edited by] I. Kelman Cohen, Robert F. Diegelmann, William J. Lindblad. Philadelphia : W.B. Saunders Co., c1992.
93	Little work has been done on the cytoskeletal mechanism of epidermal cell motility, but it is recognised that epidermal cells in all strata of the epidermis contain actin and the motile machinery probably involves the actin-myosin contractile system	Little work has been done on the cytoskeletal mechanisms of epidermal cell motility, but it is recognized that epidermal cells in all strata of the epidermis contain actin and that the motile machinery probably involves the actin-myosin contractile system.	Page 124. Wound healing : biochemical & clinical aspects / [edited by] I. Kelman Cohen, Robert F. Diegelmann, William J. Lindblad. Philadelphia : W.B. Saunders Co., c1992.
93	A cytoskeletal model of epidermal cell motility has been proposed by Bereiter-Hahn et al. (1981), in which the motive force is generated by directed contractions of the actin filaments in the cell body, forcing the cytoplasm .....	A cytoskeletal model of epidermal cell motility has been proposed by Bereiter-Hahn and associates45 in which the motive force is generated by directed contractions of the actomyosin system, forcing cytoplasm  ....	Page 124. Wound healing : biochemical & clinical aspects / [edited by] I. Kelman Cohen, Robert F. Diegelmann, William J. Lindblad. Philadelphia : W.B. Saunders Co., c1992.
93	It is generally held that a free edge is all epithelium needs to initiate movement....	It is generally held that a free edge is all epithelium needs to initiate movement.	Page 123. Wound healing : biochemical & clinical aspects / [edited by] I. Kelman Cohen, Robert F. Diegelmann, William J. Lindblad. Philadelphia : W.B. Saunders Co., c1992.
93	However, this concept may be an oversimplification since epidermal cells will not migrate in cell culture unless the substratum is optimal even though the epidermal cells have a free edge...	However, that concept may be too simple since primary epidermal cells not adapted to culture will not spread in tissue culture unless the substratum is optimal even though the cells have a free edge......	Page 124. Wound healing : biochemical & clinical aspects / [edited by] I. Kelman Cohen, Robert F. Diegelmann, William J. Lindblad. Philadelphia : W.B. Saunders Co., c1992.
93	....it appears likely that a free edge may provide the stimulus for the epithelial cells to spread, but .....	....it appears likely that a free edge may provide the external stimulus to spread, but ....	Page 124. Wound healing : biochemical & clinical aspects / [edited by] I. Kelman Cohen, Robert F. Diegelmann, William J. Lindblad. Philadelphia : W.B. Saunders Co., c1992.
84	Recent experiments support the fact that under certain circumstances, mesenchymal cells may transform to become part of the regenerating epithelium (Chong et al. 1987).	Recent observations suggest, however, that under some circumstances mesenchymal cells may transform and become part of the regenerating epithelium;26	Page 119. Wound healing : biochemical & clinical aspects / [edited by] I. Kelman Cohen, Robert F. Diegelmann, William J. Lindblad. Philadelphia : W.B. Saunders Co., c1992.

The book "Wound healing: biochemical & clinical aspects" is available in the HKU Dental Library

Plagiarism and self-plagiarism in Prof. Cheung's student's PhD thesis

Hannah Chua completed the PhD under the supervision of Prof. Lim Cheung. In a ceremony held on 14^th Dec 2010, Hannah Chua was awarded the medal for the most outstanding PhD dissertation submitted from the Faculty of Dentistry for the academic year 2008/2009. However now it has come to light that this thesis contains masive amount of plagiarism as well as self plagiarism from her MDS thesis.

In page i of her thesis Hannah Chua in the declares

I declare that this thesis represents my own work, except where due acknowledgement is made, and that it has not been previously included in a thesis, dissertation or report submitted to this University or to any other institution for a degree, diploma or other qualification.

The Regulations for The Degree of Doctor of Philosophy states that a PhD candidate's research should be "an original contribution to knowledge". If Hanah Chua is merely regurgitating what was done for her master's degree along with text copied from Dr. Nattharee Chanchareonsook's and Dr. John Ser-pheng Loh's master's  thesis as well as several other journal articles does it constitute "original contribution to knowledge"? As illustrated in the  tables below even some of the results of the old master's thesis have been self-plagiarized in the PhD. Does repeating the same results meet the criteria for an original contribution to knowledge? Does such a plagiarized dissertation warrant a PhD or the award as the most outstanding thesis?  

Academic fraud related to Hannah Chua's PhD thesis

1. PLAGIARISM

Hanah Chua has included in her PhD thesis, chunks of text copied from a Master's dissertations submitted by Dr. Nattharee Chanchareonsook to HKU in 2004 titled "Speech outcome and velopharyngeal function in Cantonese cleft patients : comparison of le fort I maxillary osteotomy and distraction osteogenesis: a pilot study" and Dr. John Ser-pheng Loh in 2004 titled “The psychological profile of cleft and non-cleft patients presenting with dento-facial deformities and its changes following surgery”. Even though Dr. Loh's thesis can be downloaded from the HKU scholar’s hub website http://hub.hku.hk/handle/10722/40702 the university has blocked access to dissertations of Dr. Chua and Dr. Chanchareonsook. 

Hannah Chua's PhD Thesis – 2008 (Won the most Outstanding Thesis Award)		Dr. Nattharee Chanchareonsook's Master's Thesis – 2004 (Did not win any awards)
Page 46	All patients included in the study received nasoendoscopy examination preoperatively and at 3, 12 and 24 months post-operatively to determine their velopharyngeal status.	Part II -Page 8	All subjects enrolled in the study received nasoendoscopy examination preoperatively and at 3 months postoperatively to determine their velopharyngeal status.
Page 46	All patients were examined using a flexible nasoendoscope (Kay Rhino-Laryngeal Stroboscope RLS 9100B, KayPentax, Lincoln Park, NJ; Olympus ENF type P-4, Olympus America Inc., Melville, NY) (Fig. 10a).	Part II- Page 8	All patients were examined with a flexible nasoendoscope (Rhino-Laryngeal Stroboscope RLS 9100B KAY (Figure B), Olympus ENF type P-4 (Figure C).
Page 46	The scope was connected to a cold light supply to allow visualization.	Part II-Page 8	The scope is connected to a cold light supply to allow visualization..............
Page 46	The image obtained from the scope was visualized on a colored monitor.....	Part II-Page 8	The image obtained from the scope was visualized on a color monitor..........
Page 46	......patients were asked to sit with their heads extended and tilted backward about 15 degrees .......	Part II-Page 9	.....patients were placed in a sitting position and their heads were extended about 15 degrees .......
Page 46	Topical anesthesia using 4% cocaine was sprayed into the nasal cavity in order to maximize patient comfort and cooperation. The nasoendoscopy was performed using the same nostril in each subject at the pre-operative and post-operative examinations	Part II-Page 9	Topical anesthesia using 4% Cocaine spray was used to maximize patient comfort and cooperation and minimize secretions. The nasoendoscopy was performed using the same nostril in each subject at the preoperative and postoperative examinations
Page 46 & 47	The boundary of the velopharyngeal orifice is defined by the posterior border of the soft palate (velum), left and right lateral pharyngeal walls and the posterior pharyngeal wall.	Part II-Page 9	Velopharyngeal orifice boundaries (the velum posterior border and velar point, left and right lateral pharyngeal walls and the posterior pharyngeal wall).............
Page 47	The "Cantonese Nasoendoscopy Speech Protocol," developed by Whitehill176 was used. The protocol included the patients' name, age, counting from 1 to 20, sustained production of two vowels, consonant vowel repetitions (e.g. "pa pa pa", "ti ti ti"), sustained /s/, sentences loaded with plosives, fricatives and nasals, and nonspeech activities such as blowing, whistling and swallowing.	Part II-Page 8 & 9	The 'Cantonese Nasoendoscopy Speech Protocol' which was developed by T. Whitehill of the Department of Speech & Hearing Sciences, University of Hong Kong (January 2000) was used in this examination. The protocol includes the client's name, age, counting from 1 to 20, sustained vowel products, consonant vowel (CV) repetitions, sustained /s/, connected speech (sentence repetition) and non-speech activities such as blowing and whistling.
Page 47	Each speech assessment took about 3-5 minutes to complete.	Part II-Page 8	Each assessment took 3-5 minutes to complete.
Page 47	For patients presenting with pharyngeal flap, both pharyngeal orifices were included in the assessment so that all borders could be seen in a single view. Each portal was also examined individually.	Part II-Page 9	For patients presenting with pharyngeal flap, both pharyngeal orifices were included so that all borders could be seen in a single view. Each portal was also examined individually.
Page 48	The clinical nasoendoscopy data were analyzed upon completion of each examination and the diagnosis was noted in the clinical record. This included: classification of velopharyngeal function (adequate closure, borderline competence, mild, moderate or severe incompetence), the consistency of the velopharyngeal function, any phoneme effects (e.g. better closure achieved for plosives), and pattern of closure (e.g. circular, coronal).	Part II-Page 9	The clinical nasoendoscopy data was analyzed by both investigators upon completion of each examination and a judgment noted in the clinical record, including classification of velopharyngeal function (adequate velopharyngeal function, borderline competence, mild, moderate or severe VPI). The consistency of velopharyngeal function was also noted.
Page 48	All the nasoendoscopy data were retrieved and transferred to a mini-digital video tape (Sony Corporation, Tokyo, Japan) by a technician not involved in the study.	Part II-Page 10	All nasoendoscopic data were transferred randomly to compact discs by a technician not associated with the study.
Page 48	The videos were then renamed by number and randomized across groups (CO and DO) and the time of evaluation (pre-op and post1 and post2).	Part II-Page 10	The digital samples were randomized across groups (conventional osteotomy and distraction), patients and time of evaluation (preoperatively or postoperatively).
Page 48	The analysis design was a randomized blind evaluation of the velopharyngeal gap closure.	Part II-Page 10	The analysis thus entailed a randomized blind estimation of the velopharyngeal gap closure.
Page 48	Table 25	Part II-Page 40	Identical to Table 1
Page 49	If the pattern of velopharyngeal closure was inconsistent, the maximum velopharyngeal closure during the entire protocol was used to represent the final rating.	Part II-Page 10	If the pattern of velopharyngeal closure was inconsistent, the maximum velopharyngeal closure during the entire protocol was used to represent the final rating.
Page 49	Speech assessment Speech examinations were carried out on the same visit as the velopharyngeal assessment at the Division of Speech and Hearing Sciences, University of Hong Kong.	Part II-Page 12	Speech Examination Speech examination was carried out preoperatively and 3 months postoperatively for each patient at the Division of Speech and Hearing Sciences, University of Hong Kong.
Page 49	During each session, an experienced speech-language pathologist examined each patient to evaluate resonance (hypernasality and hyponasality) nasal emission, and articulation.	Part II-Page 12	During each session, an experienced speech language pathologist (a doctoral student) assessed the patient for hypernasality, nasal emission, and articulation.
Page 49	The speech examination was conducted in a quiet room.	Part II-Page 12	The examination was conducted in a quiet room.
Page 49	Each patient was asked to read a set of Cantonese/Chinese speech materials including (a) a written passage ("Barbara Streisand" passage), (b) counting from 1-20, (c) the Cantonese Osteotomy Deep Test, a 120-item single-word test, loaded with phonemes known to be vulnerable in speakers with Class III malocclusion.179 -180	Part II-Page 12	The speech sample for evaluation of hypernasality and nasal emission consisted of Barbara Streisand passage. The speech sample for the articulation consisted of the Cantonese Osteotomy Deep Test developed by T. Whitehill of the Department of Speech & Hearing Sciences, University of Hong Kong (Whitehill et al, 2000)
Page 49	The two tests above and the findings were recorded using a Sony TCD-D3 Digital (DAT) tape recorder. A Sony ECM-909 microphone was used, with a mouth to microphone distance maintained at approximately 10 cm. The speech assessment process was also recorded using a JVC GR-AX7E video camera positioned to allow maximum view of the mouth during recording.	Part II-Page 12	Speech samples were recorded using a Sony TCD-D3 Digital (DAT) tape recorder. A Sony ECM-909 microphone was used and a mouth to microphone distance of 10 cm was maintained. The speech assessment was also recorded using a JVC GR-AX7E video camera positioned to allow maximum view of the mouth during recording.
Page 50	The audio recordings were renamed as numbers, randomized (across groups and times of evaluation)	Part II-Page 10	The digital samples were randomized across groups (conventional osteotomy and distraction), patients and time of evaluation (preoperatively or postoperatively).
Page 50	Table 26	Part II-Page 41	Identical to Table 2
Page 51	Nasality was assessed using a Nasometer (Kay Elemetrics, Model 6200, KayPENTAX, New Jersey, USA). The Nasometer was calibrated according to the manufacturer's instructions, and the headset device was adjusted by the speech language pathologist ....	Part II-Page 15	Key Elemetrics Model 6200 Nasometer (Figure F) was used for data collection. The Nasometer was calibrated, and the headgear was adjusted by the examiner, a qualified speech language pathologist, according to the manufacturer's instructions.
Page 51	This headset device separated the oral and nasal cavities using a baffle plate. Two microphones mounted on the top and bottom of the plate collected acoustic energy during speech.  The instrument then computes a ratio of the acoustic data acquired by the two microphones (nasal energy divided by oral plus nasal energy) (Fig. 14b). This ratio is called nasalance, which is considered an acoustic correlate of perceived nasality and is expressed in percentage	Part II-Page 15	The headset consists of two microphones, one on either side of a metal plate that rests horizontally on the upper lip. The sound signal from each microphone is transmitted to the computerbased nasometer. The instrument measures the relative amount of nasal acoustic energy in a subject's speech in the form of "nasalance" which is a ratio of nasal acoustic energy divided by nasal plus oral acoustic energy, expressed as a percentage.
Page 51	The calibration was made before assessment of each subject. Each cleft patient read aloud an oral passage,181 which contained no nasal phonemes. If an error occurred during reading, the subject was asked to repeat the test.	Part II-Page 15	The calibration was made before assessment in each subject. Subjects read the speech stimuli aloud. If an error occurred during reading, the subject was asked to repeat the stimulus.
Page 51	The mean percentage nasalance was calculated for each cleft patient at each time period. Previous studies have recommended various cut-off points for nasalance values such as 32% and 26%. In this study, a nasalance value higher than 30% was considered to be indicative of hypernasality. A 2-point nasalance scale was thus derived, 1 = above/outside normal limits (nasalance above 30%), 0 = within normal limits (30% nasalance or below).	Part II-Page 16	The mean percentage nasalance was collected for each speaker. The previous studies reported the cutoff point nasalance values; for example 32% (Dalston et al, 1991) and 26% (Hardin et al., 1992. Based on the average of these two values, nasalance values higher than 30%, were considered to be suggestive of increased hypernasality. A 2-point nasalance scale was thus derived, 1= Hypernasality (nasalance above 30%), 0= No hypernasality (30% nasalance or below).
86	The nasoendoscopy results were further classified into patients who showed an improvement, no change, and deterioration in velopharyngeal status according to the types of surgery (Table 46). An Exact Chi-square test showed no significant difference between the number of subjects who showed "improvement", "deterioration", and "no change" in their velopharyngeal status between the 2 types of surgery.	Part II- Page 21	The nasoendoscopy results were classified into subjects who showed an improvement in velopharyngeal status, subjects who showed no change, and those who showed deterioration in velopharyngeal status. Table 6 shows the distribution of these 3 categories across the twu surgical groups. An Exact Chi-square test showed no significant difference between number of subjects who showed ''improvement'', "deterioration", and ''no change" of the velopharyngeal status between the 2 types of surgery.
97	...were classified into those who "deteriorated", "improved" or had "no change". The distributions of the 3 categories across the two surgical groups are presented in Table 52. The Chi-square test showed no significant difference between number of patients who showed "improvement", "deterioration", and "no change" of the velopharyngeal status between the 2 types of surgery.	Part II- Page 21	......were classified into subjects who showed an improvement in velopharyngeal status, subjects who showed no change, and those who showed deterioration in velopharyngeal status. Table 6 shows the distribution of these 3 categories across the two surgical groups. An Exact Chi-square test showed no significant difference between number of subjects who showed ''improvement'', "deterioration", and ''no change" of the velopharyngeal status between the 2 types of surgery.

The similarities between Dr. Chua's and Dr. Loh's dissertations can be found here.

In addition to the above, Hannah Chua plagiarized text from several journal articles. The link to the original source can be found at the last column.

Page No.	Plagiarized text in Hannah Chua's PhD Thesis – 2008	Original Text	Original Source
Page 27	Apart from the stability and relapse, complications and impact on speech and velopharyngeal function, aesthetics, patients' satisfaction with treatment and impact of treatment on social functioning need to be further evaluated.	Apart from the surgical aspects, long-term stability and relapse, complications and impact on speech and facial aesthetics, patient satisfaction with treatment, and impact of treatment on social functioning need to be further evaluated.	Bas van Eggermont, J. Jansma, M.W.J. Bierman, B. Patient satisfaction related to rigid external distraction osteogenesis. International Journal of Oral & Maxillofacial Surgery. (Volume 36, Issue 10 , Pages 896-899, October 2007)
Page 26	Maxillary distraction osteogenesis is believed to be effective for facial and occlusal improvement, especially in CLP patients.	Maxillary distraction osteogenesis is believed to be effective for facial and occlusal improvement, especially in cleft lip and palate patients......	Kiyoshi Harada, Yoshiyuki Baba, , Kimie Ohyama, , Ken Omura. Soft tissue profile changes of the midface in patients with cleft lip and palate following maxillary distraction osteogenesis: A preliminary study. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology and Endodontology (Volume 94, Issue 6 , Pages 673-677, December 2002)
Page 26	However, few reports have described the soft tissue profile changes after maxillary distraction.	However, few reports have fully described changes in the soft tissue profile in cleft lip and palate patients after maxillary distraction
Page 22	It has been suggested that the membranous bone of the craniofacial skeleton may be at risk of premature union if the distraction rate is started too slow.	It has been suggested that the membranous bone of the craniofacial skeleton behaves differently and that there is a risk for premature consolidation if the distraction rate is too slow,	Swennen G, Schliephake H, Dempf R, Schierle H, Malevez C. Craniofacial distraction osteogenesis: a review of the literature: Part 1: clinical studies. Int J Oral Maxillofac Surg. 2001 Apr;30(2):89-103.
Page 22	According to Ilizarov, the success of distraction osteogenesis depends on the response of the initial callus to tensile strength.	According to ILIZAROV54,55 success of DO depends on the response of the initial callus to tensile stress.
Page 22	A latency period of 5-7 days is therefore generally recommended for endochondral bone to allow time for initial callus formation and healing of soft tissues. The membranous bones of the craniofacial skeleton, in particular the maxilla, have a rich blood supply,....	.....a latency period of 5-7 days after surgery is necessary to allow time for initial callus formation and healing of soft tissues54 56. The membranous bones of the craniofacial skeleton, which are thin and have a rich blood supply....
Page 23	After an active period of distraction, the distractors are left in place to allow adequate consolidation and maturation of the bone callus. This review showed that....	After the active period of distraction, the distraction appliances are left in place for adequate consolidation and maturation of the bony callus. This review revealed that.......
Page 23	consideration the type of skeletal deformities and the patients need	....consideration to the type of skeletal deficiency and the patient's needs..
Page 122	Tindlund and Rygh197 found that after orthopedic maxillary advancement, patients with bilateral CLP showed more nose and upper lip advancement and more backward movement of the lower lip.	........Tindlund and Rygh10 found that after orthopedic maxillary advancement, patients with bilateral CLP showed more nose and upper lip advancement and more backward movement of the lower lip.......	Wen-Ching Ko E, Figueroa AA, Polley JW. Soft tissue profile changes after maxillary advancement with distraction osteogenesis by use of a rigid external distraction device: a 1-year follow-up. J Oral Maxillofac Surg. 2000 Sep;58(9):959-69.
Page 122	.......they concluded that the overall soft tissue profile changes were similar in both bilateral and unilateral CLP groups.	......they concluded that the overall soft tissue profile changes were nearly the same in the bilateral and unilateral groups.
Page 25	...distraction osteogenesis is successfully applied to correct maxillary hypoplasia.....	Distraction osteogenesis has been successfully applied to correct midface hypoplasia.....
Page 122	.....the small sample size in each cleft type...	..... the small sample size in each cleft type,

Pages 22, 23 & 25 of Hannah Chua's PhD are common to her master's thesis as well (see below). This means part of her Master's dissertation was also plagiarized.

2. SELF-PLAGIARISM

The entire Chapter 1 (page 1-27) of the PhD thesis and some parts of Chapter 3-5 are an almost exact word by  word reproduction of Hannah Chua's own Master's thesis submitted to HKU in 2004 titled "Cleft Maxillary Distraction versus Orthognathic Surgery: Clinical Morbidities and Surgical Relapse". 

Hannah Chua's PhD Thesis – 2008 (Won the most Outstanding Thesis Award)			Hannah Chua's Master of Dental Surgery Thesis - 2004
Section	Pages	Text in the PhD	Pages	Identical text in the Masters' thesis
Chapter 1 - Literature Review	1-27	I do not even bother to make a side-by-side comparison for this chapter due to space constraints.	2 - 19	The similarity Chapter 1 of the PhD and Part 1 of the master's thesis is so obvious (apart from some minor modifications in the tables and numebrs)
Chapter 3 - Methods	34	The mobilized maxilla was fixed with 2 titanium mini-plates on each side at the zygomatic buttress and the pyriform region	52	The mobilized maxilla was fixed with 2 titanium mini-plates on each side at the zygomatic buttress and the pyriform region
	36	After a latency of 3 days, activation was commenced at 1 mm per day in 2 rhythms with an activating key by the patients themselves or their relatives until a class I incisal relationship was achieved.	52	After a latency of 3 days, activation was commenced at 1 mm per day in 2 rhythms until a class I incisal relationship was achieved.
	36	Light orthodontic elastic was applied to control the occlusion in both groups during the early post-operative period.	52	Light orthodontic traction was used to control the occlusion in the early post-operative period in both groups.
	37	Standardized questionnaires were used to record the clinical morbidities during the operation and post-operatively.	52	Standardized questionnaires were used to record the clinical morbidities during the operation and post-operatively.
	37	The patients were reviewed at regular postoperative intervals	53	The patients were reviewed at regular post-operative intervals
	37	Standardized lateral cephalographs were taken on all patients enrolled in this study both shortly before the operation and post-operatively	53	Standardized lateral cephalographs were taken of all patients enrolled in this study both shortly before the operation and post-operatively
	37	All tracings were superimposed using the sella, nasion and cranial base structures employing a method of anatomic best fit.	53	All tracings were superimposed using the sella (S), nasion (N) and cranial base structures employing a method of anatomic best fit.
	37	The linear movements of the landmarks were measured using an electronic digital caliper (Digit Cal, Tesa, Switzerland) with accuracy up to 2 decimal points.	54	The movements of the landmarks were measured using an electronic digital caliper (Digit Cal, Tesa, Switzerland) down to 2 decimal points.
	37	Each reading was taken three times and a mean value was recorded	54	Each reading was taken three times and the mean was documented.
	43	40 lateral cephalographs randomly selected from 20 patients.	54	40 lateral cephalographs were randomly selected from 20 patients.
	43	The same measurement was repeated one week later. Reliability and random error analyses on the 2 sets of measurement were performed. The reliability of the measurements were evaluated by paired t-test with a 5% level of significance (SPSS Incorporations, Chicago, IL, USA). The reproducibility or random error is of importance since it determines the metric differences between the two measurements and the acceptability of the differences. Dahlberg's formula175 was used to determine the random error.	54	the same process was repeated by the same author one week later. Reliability and error analysis tests from the 2 sets of measurement were performed. Reliability was confirmed by paired t-test with a 5% level of significance (SPSS, Chicago, USA). The random error is of relevance since it determines the metric differences between the two measurements and the acceptability of the differences. Dahlberg's formula23 was used to determine the random error.
	43	The reliability analysis using paired t-tests showed no significant difference between two tracings carried out at separate occasions	54	The reliability analysis using paired t-tests showed no significant difference between two tracings carried out at separate occasions.
	44	The random error analysis confirmed a small difference, but it was within the clinically acceptable limit	54	The random error confirmed a small difference, but it was within a clinically acceptable limit
Chapter 4 - Results	59	In the CO group, one patient (C10) experienced rupture of the right descending palatine vessel during the operation (Table 28). The vessel was subsequently ligated to control the bleeding. Post-operatively, the patient was closely monitored and the healing was uneventful.	56	In the osteotomy group (Table 2), one patient experienced rupture of the right descending palatine vessel during the operation. The vessel was subsequently ligated to control the bleeding. Post-operatively, the patient was closely monitored and the healing was uneventful.
	60	In the CO group, one patient (C6) developed early clinical relapse of dental occlusion to Class III edge to edge three months after the surgery. The occlusion was later compensated by orthodontics	56	One patient developed early clinical relapse of dental occlusion to Class III edge to edge three months after the surgery. The occlusion was later compensated by orthodontics.
	60	Another patient (C7) had exposure of one titanium screw at the buccal sulcus region and the condition was complicated by maxillary sinusitis (Fig. 21a&b). He received antrostomy and removal of the plates and screws from the affected site, and the complications were subsequently resolved.	56	Another patient had exposure of one mini-plate at the buccal sulcus region and the condition was also complicated by maxillary sinusitis (Fig 4). The patient received antrostomy and removal of the plates and screws of the affected site at the same operation and the complications were subsequently resolved.
	59	One patient (D1) developed early clinical relapse of dental occlusion to edge to edge occlusion 7 weeks after distraction surgery. This malocclusion was attempted to be corrected by transposing the maxilla using a Le Fort I osteotomy to the planned class I occlusion and fixation by titanium mini-plates and screws at the distractor removal stage. However, at 10 weeks after this second surgery, the occlusion relapsed again to class III malocclusion resulting in an overjet of -2.5mm. Compensation by orthodontics was tried but the occlusal result was still consideredunsatisfactory (Fig. 18a-c).	56	One patient developed early clinical relapse of dental occlusion to class III malocclusion (overjet = -4mm) two m onths after distraction surgery. This malocclusion was corrected by a repeat Le Fort I osteotomy and fixation by titanium mini-plates and screws at the distractor removal stage. However, at eight weeks after this second surgery, the occlusion relapsed again to class III malocclusion (overj'et = - 2.5mm). Compensation by orthodontics was tried but the occlusal result was still considered unsatisfactory (overjet = -1mm, open bite = 1mm).
	55	Figure 15	55	Similar to Figure 5
	62	Figure 21	69	Similar to Figure 6
Chapter 5 - Discussion	113	On the other hand, surgical advancement of <4mm can be achieved by cleft orthognathic surgery relatively easily. The choice between orthognathic surgery and distraction largely depends on the extent of maxillary advancement required. This study aims to provide evidence-based data in clarifying the advantages and disadvantages of each treatment method	57	On the other hand, small surgical advancement of <4mm can be achieved by cleft o rthognathic surgery relatively easily. The choice between orthognathic surgery or distraction largely depends on whether the patients need moderate advancement of the maxilla, and this study aims to provide evidence-based data in clarifying the advantages and disadvantages of each treatment method.
	113	Different techniques of maxillary distraction osteogenesis have been developed, depending on the type of devices used. Distraction was initially applied to the maxillofacial region with the use of external distractors.	57	Different techniques of maxillary distraction osteogenesis have been developed according to the distractor evolution, and depend on the types of devices used. Distraction was initially applied to the maxillofacial region with the use of extra-oral devices.
	114	The use of transcutaneous pins and wires can lead to traction markings that may leave lifetime residual scars on the exposed areas of the midface.	57	The use of transcutaneous pins and the traction markings may leave lifetime residual scarring.
	115	both insertion and removal require a separate operation and the retention of the device is required over an extended period for consolidation. This in turn lengthens the distraction treatment. The activation arm of the device tends to protrude into the oral cavity. The extension of an activating rod in the buccal sulcus can produce significant discomfort for patients. Oral hygiene maintenance and intake of food cause patients some inconvenience as well.	57 & 58	Both the insertion and removal requires operation and the retention of the device over a three month period. This in turn lengthens the distraction treatment The activation arm of the device tends to protrude into the oral cavity. This can produce significant discomfort for the patients. Oral hygiene maintenance and intake of food also cause patients some inconvenience.
	115	Although distraction is expanding its role in the maxillary transverse widening with the use of palatal distractors, it cannot yet achieve three or four pieces of segmentalization like orthognathic surgery. Retrusion or impaction of the maxilla is also only possible at present by orthognathic surgery.	58	Although distraction is expanding its role in the maxillary transverse widening with the use of palatal distraction, it cannot yet achieve the three or four pieces of segmentalization like orthognathic surgery. Distraction also cannot achieve retrusion or impaction of the maxilla that is currently possible only by orthognathic surgery.
	115	There have been few reports focusing on the complications from distraction osteogenesis. A similar range of morbidities and complications associated with conventional orthognathic procedures may also occur with distraction osteogenesis.	58	There have been few reports on the potential complications concerning distraction osteogenesis. A similar range of morbidities and complications associated with conventional orthognathic procedures may also occur with distraction osteogenesis.
	116	Since the rod penetrates the mucosa in the mucobuccal fold, it is a portal of entry for microorganisms predisposing to infection.	58	Since the rod penetrates the mucosa in the mucobuccal fold, it is a portal of entry for microorganism predisposing to infection.
	116	Good hygiene is hence very important for patients fixed with internal distractors.	58	Good hygiene is hence very important for patients having distraction via an intra-oral approach.
	116	In a study performed by Kebler,196 four patients also developed mucosal infection at various times during the distraction process, which was satisfactorily controlled by antibiotics.	59	In a study performed by Kepler26, four patients also developed mucosal infection at various times during the distraction process, which was satisfactory controlled by antibiotics.

3. SUPERVISORS

In page ii of the PhD thesis, Hannah Chua wrote the following hosanna "I would like to gratefully acknowledge the constant and invaluable academic and personal support received from my supervisor, Professor Lim K. Cheung. Without his dedicated supervision, it would have been difficult to complete this thesis on time. He has indeed spent a huge amount of time and energy in guiding me through the writing of this thesis, correcting and revising the text that is about to be read. His insightful and valuable feedbacks, as well as suggested improvements have transformed this thesis into a more concise and readable form" (emphasis mine)

If Prof. Cheung "spent a huge amount of time and energy" guiding Hannah Chua on how to write the thesis and "correcting and revising the text" he should also be held responsible for the plagiarism. In addition, Prof.Cheung was the supervisor for Hannah Chua's master's thesis. Therefore he would have known she was submitting the same text in her PhD.

4. EXAMINERS

All PhD students at HKU have a supervisor and one or more co-supervisors. When appointing examiners for the PhD, Prof. Cheung recommended one of Hannah Chua's co-supervisor's spouse as an examiner. This recommendation was approved by the faculty higher degrees committee headed by Prof. Cheung.  

HKU's guidelines to supervisors on appointment of examiners state "For the student's own sake, any situation potentially giving rise to a conflict of interest should be avoided, and be seen to be avoided" (Clause w)

In order not to compromise the examination process and to enable it to be conducted in a fair manner, neither you nor your students should communicate with the External Examiners during the examination process until the Board of Graduate Studies (BoGS) has made a final decision on the examination. (Clause x)

5. POTENTIAL DATA FALSIFICATION

In the Master's Thesis

Since Hannah Chua 1st submitted her master's thesis, there was a lot of controversy in the Faculty of Dentistry that Part I of her thesis was not a true meta-analysis. The main arguments were that putting together a large number of case reports could not constitute a meta-analysis. Another concern raised by many staff members (in private) were that the examiner appointed to evaluate Hannah Chua's master's thesis Dr. Chantal Malavez has never published a meta-analysis by herself (a list of Dr.Malavez publications can be found here).  It is an accepted tradition that a person who is examining a specialized theme (such as a meta-analysis) should have published at least one paper on the particular topic.

Now Prof. Cheung and Hannah Chua appear to have accepted the concerns of their colleagues because, even though reproducing the same text of her master's thesis as Chapter I of the PhD, they refrain from labelling it as a meta-analysis. They simply call it a comprehensive literature review. The exact wordings are as follows;

Master's thesis (page 4)

A considerable body of literature now exists on the application of both conventional osteotomy and distraction osteogenesis to CLP patients. The aim of this study is to conduct a comprehensive meta-analysis of this literature, and compare the results achieved by these two forms of surgical treatment.

PhD thesis (page 2)

A considerable body of literature now exists on the application of both conventional osteotomy and distraction osteogenesis to CLP patients. The aim of this comprehensive review of the literature is to compare the clinical treatment outcomes of cleft maxillary distraction and cleft orthognathic surgery.

Possible Data Falsification in the PhD Thesis

In response to regurgitating some of the results from the master's thesis, I am sure Prof. Cheung will argue that this was a longitudinal study and therefore they have to follow up the same patients.

But In page 56 of the master's thesis, it is stated "In the distraction group (Table 1), two patients developed mucosal infection around the distractors during the activation period. The infection was controlled by intravenous antibiotics administered for a week." But in page 59 of the PhD thesis it is stated "In the distraction group (Table 27), 10 patients developed mucosal infection around the distractors during the consolidation period (Fig 17). The infection was controlled by oral antibiotics administered for a week."

The activation (during which the device called the "distractoion rod" is rotated) and consolidation periods (when no active interventions are done) are 2 entirely different time points. Similarly, oral and intravenous are 2 entirely different methods of administering antibiotics. So what happened to theses 2 patients included in the master's thesis; if Hannah Chua was following up the same patients. How did they disappear from the PhD thesis? This is an indication of potential data fabrication.