A 5-year-old girl of African descent presented with a history of progressive painless swelling on the right side of the jaw since the past 2–3 months. Orthopantomogram showed a radiolucent lesion near the angle of the mandible. Subsequent CT scan revealed a 2 cm×2 cm radiolucent lesion with intense periosteal reaction surrounding the lesion and destruction of the overlying cortex. Radiological perplexity aroused regarding the possibility of eosinophilic granuloma or some other malignant lesion. Incisional biopsy performed and microscopy showed spindle cell tumor. Immunohistochemistry confirmed it as myofibroma. Myofibroma is a rare benign tumour involving mesenchyme. Involvement of the mandible is rare. Radiological presentation with strong periosteal reaction is a rarity and has rarely been reported in the medical literature. We conclude that intraosseous myofibroma can sometimes have strong periosteal reaction and careful radiological evaluation is a prerequisite for accurate diagnosis and to avoid unnecessary aggressive therapy.

We read with interest the recent article by Cibulskis and Armbrecht 1 describing the association of metabolic acidosis in premature infants fed on breast milk fortified with liquid fortifier. As our neonatal unit has reached 100% breast milk exposure to extremely low birth weight (ELBW) and 90% for very low birth weight (VLBW) infants through a quality improvement project, 2 we are keen on the author's observations. However, we wish to share some reservations regarding the methodology adopted to arrive at the associations and conclusions:(1) There is considerable room for selection bias between the two groups being compared. Two consecutive patient populations were selected at different time periods by block allocation. 3 A number of potential confounders, that could have influenced the study results, were left to the treating clinician's discretion. Generally a good randomisation process achieves groups that are similar in baseline characteristics, and thereby minimises confounding and type-1 errors that could lead to misleading results. 4 (2) Sufficient information is not provided on the acid-base status of the infants prior to administration of the liquid human milk fortifier (HMF). We also have no visibility in the paper of some key co-morbidities or pharmaceutical agents that might have contributed to metabolic acidosis, for example, sepsis, significant patent ductus arteriosus or regular medications with acidic pH such as caffeine citrate. (3) No institutional feeding guidelines have been explained or given as an appendix in the paper, and no clear values proposed advising clinicians regarding at what level of base excess/pH the HMF should be discontinued. (4) There is no logistical regression done and as such no association could be precisely deduced between metabolic acidosis and use of the HMF. When one nominal variable (dependent) and one measurement variable (independent) are in question, we would like to know whether variation in the measurement variable causes the variation in nominal variable. We do recognise that the authors did not attempt a causation hypothesis; however, the overall methodology permits only a weak association at the best.In our unit we use traditional powdered HMF and have not observed metabolic acidosis attributable to or associated with any of our ELBW (30) or VLBW (55) infants whose acid base status was prospectively collected as an additional variable while collating the Vermont-Oxford Network (VON) data for 2013 and 2014. Table 1 compares the contents of the powdered HMF we use (product C) with those used in the study. The compositions of the HMF described in the paper and ours are similar. We used mother's own breast milk as well as donor human milk from the bank during the period of observation. We also have a very low necrotising enterocolitis (NEC) rate-no cases in 2013 (with an inhouse birth rate of 4265), that too has been achieved without additional probiotic use. We have no unit-level experience with the liquid HMF and are hence unable to compare our results with the...

Conventional radiography remains a key diagnostic tool in neonatology particularly in VLBW and ELBW infants and is invaluable in supporting timely clinical decision making. Clinicians should be aware of the cost and potential hazards of neonatal radiography and is recommend that the cumulative radiation exposure among the ELBW and VLBW infants is monitored. Increasing awareness and standardisation of point-of-care ultrasonography could decrease the reliance on conventional radiography in neonatal units.