This case, which is subject to the Texas Medical Liability Act (TMLA), Tex. Civ. Prac. & Rem. Code ch. 74, involves the adequacy of an expert report. The issue is whether the trial court abused its discretion by denying the defendants’ motion to dismiss in light of conflicting statements in the plaintiffs’ expert report, some of which the defendants alleged, and the court of appeals held, failed to link the expert’s conclusions to the underlying facts.
Nicholas Van Ness died from pertussis (whooping cough) when he was two months old. His parents, Melissa and Ronald Van Ness, sued Kristin Ault, D.O., and her employer, ETMC First Physicians, alleging that Dr. Ault’s negligence caused Nicholas’s death and that ETMC was vicariously liable for her negligence. The Van Nesses timely served Dr. Ault and ETMC with an expert report by Alvin Jaffee, M.D., then served an amended report after the trial court sustained the defendants’ objections to the original. The defendants again moved to dismiss the suit, contending that Dr. Jaffee’s opinions as to causation were conclusory because the amended report (the report) failed to link his opinions to the underlying facts. The trial court denied the motion. On interlocutory appeal, see Tex. Civ. Prac. & Rem. Code § 51.014(a)(10), the court of appeals reversed and ordered the suit dismissed. ETMC First Physicians v. Van Ness, 461 S.W.3d 152 (Tex.App.-Tyler 2014). We reverse the judgment of the court of appeals.
A plaintiff asserting a health care liability claim must serve each defendant with an expert report that includes “a fair summary of the expert’s opinions ... regarding applicable standards of care, the manner in which the care rendered by the physician or health care provider failed to meet the- standards, and the causal relationship between that failure and the injury, harm, or damage claimed.” Tex. Civ. Prac. & Rem. Code § 74.351(r)(6). A challenge to the sufficiency of a report must be sustained if “the report does not represent an objective good faith effort to comply with the [statutory requirements].” Id. § 74.351(l). A report is a good faith effort if it provides adequate information to “inform the defendant of the specific conduct the plaintiff has called into question, ... provide[s] a basis for the trial court to conclude that the claims have merit,” Bowie Mem’l Hosp. v. Wright, 79 S.W.3d 48, 52 (Tex.2002) (per curiam), and “does not con*142tain a material deficiency,” Samlowski v. Wooten, 332 S.W.3d 404, 410 (Tex.2011).
A trial court’s ruling on the sufficiency of an expert’s report is reviewed for abuse of discretion. Rosemond v. Al-Lahiq, 331 S.W.3d 764, 766 (Tex.2011); Am. Transitional Care Ctrs. of Tex., Inc. v. Palacios, 46 S.W.3d 873, 877 (Tex.2001). Under that standard, appellate courts defer to the trial court’s factual determinations if they are supported by evidence, but review its legal determinations de novo. See Stockton v. Offenbach, 336 S.W.3d 610, 615 (Tex.2011). A trial court abuses its discretion if it rules without reference to guiding rules or principles. Samlowski, 332 S.W.3d at 410. An expert must explain, based on facts set out in the report, how and why the breach caused the injury. See Jelinek v. Casas, 328 S.W.3d 526, 539-40 (Tex.2010). A bare expert opinion that the breach caused the injury will not suffice. Id.
Dr. Jaffee set out the following facts in his report as those on which he based his opinions. Nicholas was born on November 13, 2009. He was seen by Dr. Ault on November 19 and November 30 for regular checkups, and Dr. Ault noted no concerns at either visit. However, the records from his four-week checkup on December 11 reflected that Nicholas had a fever with a temperature of 100.2 degrees, was coughing and suffering from nasal congestion, and was exposed to “sick contacts at home.” According to an affidavit submitted by Nicholas’s mother, she reported to Dr. Ault on December 11 that Nicholas had been coughing to the point that he could not breathe and was exhibiting facial discoloration. Nothing indicated that Dr. Ault performed any laboratory or diagnostic tests on Nicholas.
The Van Nesses returned to see Dr. Ault on December 15 and explained that Nicholas’s symptoms had worsened. Dr. Ault physically examined Nicholas, but again did not perform or order any tests. On December 20, the Van Nesses took Nicholas to East Texas Medical Center Hospital in Jacksonville, where he was treated for acute pneumonia, wheezing, and tachycardia. The following day he was transferred to the Children’s Medical Center Hospital in Dallas, where he died on January 20, 2010.
The defendants objected to Dr. Jaffee’s report on the ground that it failed to explain how Dr. Ault’s alleged negligence caused Nicholas’s death, specifically contending that Dr. Jaffee’s medical conclusion was not linked to the facts of the case and was conclusory. The defendants moved for dismissal of the suit. The trial court denied the motion. The court of appeals reversed and rendered judgment dismissing the Van Nesses’ suit with prejudice, agreeing with the defendants that Dr. Jaffee’s report was deficient as to the causation element. 461 S.W.3d at 143.
Dr. Jaffee’s nine-page report generally discusses pertussis, including its diagnosis and treatment. His report also contains separate sections addressing the applicable standard of care, breach of the standard, and causation. In the standard of care section, he opined, in part, that
[t]he applicable standard of care as to Kristin Ault, DO is upon evaluation of a one month old child who presents with symptoms such as a history of fever, cough and nasal congestion, compounded by sick contacts at home, is to perform laboratory tests, administer antibiotics prophylaetically while the tests are pending and/or to admit the infant to a medical facility....
[[Image here]]
... [H]ad Dr. Ault, performed any of these tests, it would have shown Borde-tella pertussis at a treatable stage and but for the failure to treat Nicholas Van *143Ness as outlined above he would have had a 51% or more chance of survival.
In the breach section, Dr. Jaffee states again that Dr. Ault breached the standard of care in several ways on both December 11 and December 15, including failing to have various laboratory diagnostic tests performed on Nicholas and failing to administer antibiotics prophylactieally while the tests were performed. Finally, in the causation section of his report, Dr. Jaffee repeats his opinion that Dr. Ault should have taken specified actions including running diagnostic tests on Nicholas and administering antibiotics prophylactieally. He also states, “It is within a reasonable degree of medical certainty and/or with 51% certainty that had Dr. Ault appropriately evaluated and diagnosed him, Nicholas Van Ness would have received the appropriate dosage and treatment of antibiotics in a timely manner, and he would not have expired on January 20, 2010.”
As support for its conclusion that the report was deficient as to causation, the appeals court focused on several statements from the report, including the following:
• “prevention of pertussis via vaccination is of primary importance because treatment is of little benefit to the person infected”;
• “unvaccinated or incompletely vaccinated infants younger than 12 months of age have the highest risk for severe and life-threatening complications and death”;
• “A reasonable guideline is to treat ... infants aged less than [one] year within six weeks of cough onset”; and
• “antibiotics may ‘shorten the duration of infectiousness and are thus recommended.’ ”
461 S.W.3d 152.
In reaching its conclusion as to Dr. Jaf-fee’s report, the court of appeals noted that Nicholas began receiving treatment on December 20 in the Jacksonville hospital. It further identified December 20 as being approximately three weeks after Nicholas’s visit with Dr. Ault on November 30, at which time he was not yet feverish, coughing, or presenting other symptoms. Given the temporal sequence of events, the appeals court reasoned that treatment beginning on December 20 “was well within the reasonable guideline of treating the disease within six weeks of cough onset” specified by Dr. Jaffee and
Dr. Jaffee’s conclusion that Nicholas would not have died had Dr. Ault began treatment on December 11, 2010, or December 15, 2010, does not follow from the aforementioned discussion of the facts in his report.
Rather, because treatment is of little benefit to the person affected, the facts lead to the conclusion that had Dr. Ault provided antibiotics to Nicholas on either of those dates, at most Nicholas’s symptoms may have lessened and his ability to spread the disease to others may have diminished. The facts discussed in the report do not show that treatment would have altered the course of the disease, but lead to the conclusion that Nicholas was unfortunately one of those infants who did not survive despite timely treatment.
461 S.W.3d at 144.
The court of appeals focused on Dr. Jaffee’s statements that treatment with antibiotics is of little benefit and only would have lessened Nicholas’s symptoms and reduced the contagiousness of his pertussis. 461 S.W.3d 152. Based on those statements, the court concluded that the report showed that treatment earlier than December 20 would not have prevented Nicholas’s death; thus, the report did not *144demonstrate a causal relationship between Dr. Ault’s alleged negligence and Nicholas’s death. 461 S.W.3d 152.
In its analysis however, the appeals court did not fully credit all of Dr. Jaffee’s factual statements and opinions. In particular, the court did not credit statements and opinions from Dr. Jaffee’s report to the effect that (1) a stage existed at which pertussis could be treated with antibiotics; (2) if Dr. Ault had given Nicholas antibiotics prophylactically and ordered testing, the tests would have shown his pertussis was at a treatable stage; and (3) Nicholas would have had a 51% chance of recovery if Dr. Ault had started Nicholas on prophylactic antibiotics and continued antibiotics as indicated by results of the tests.
Dr. Jaffee’s statement about antibiotics having little effect on pertussis other than reducing the potential for spreading the disease is in tension with his statements that Nicholas was treatable with antibiotics and would have had a 51% chance of survival if Dr. Ault had administered them. The first-referenced statement, by itself, indicates that whatever Dr. Ault did on December 11 or December 15 would have had little effect on the course of Nicholas’s illness and would not have prevented his death. Accordingly, as the court of appeals explained, that statement standing alone would not demonstrate that under the facts as set out in the report Dr. Ault’s alleged negligence was causally related to Nicholas’s death, and his causation opinion would be conclusory.
However, Dr. Jaffee’s report also set out that Nicholas’s illness was treatable when Dr. Ault saw him in December, and starting antibiotics at that time and continuing them as indicated by diagnostic testing probably would have prevented Nicholas’s death. Given these parts of the report, the trial court could have determined that Dr. Jaffee’s opinions were linked to the underlying facts and explained why and how Dr. Ault’s timely treatment of Nicholas with antibiotics would have prevented his death.
Under the circumstances, the trial court had discretion — indeed it was incumbent on the trial court — to review the report, sort out its contents, resolve any inconsistencies in it, and decide whether the report demonstrated a good faith effort to show that the Van Nesses’ claims had merit. Considering both the report’s explication of how Dr. Ault’s alleged negligence was causally related to Nicholas’s death and the conflicting statements as to that causal relationship, we conclude that the trial court did not abuse its discretion by determining that the report was not conclusory, but was a good faith effort to comply with the TMLA’s requirements. Thus the trial court did not abuse its discretion by denying Dr. Ault’s motion to dismiss, see Samlowski, 332 S.W.3d at 410, and the court of appeals erred by reversing the trial court’s judgment.
We grant the petition for review. Without hearing oral argument, see Tex. R. App. P. 59.1, we reverse the court of appeals’ judgment and remand the case to the trial court for further proceedings.